Digital Radio Technology and Applications

Home , Bell 202 modem, CCIR 476, Radioteletype, Shaft (company), Telebit, Terminal node controller

DIGITAL RADIO TECHNOLOGY AND APPLICATIONS

Proceedings of an International Workshop organized by the International Development Research Centre, Volunteers in Technical Assistance, and United Nations University, held in Nairobi, Kenya, 24-26 August 1992

Edited by

Harun Baiya (VITA, Kenya) David Balson (IDRC, Canada) Gary Garriott (VITA, USA) 1 1 X 1594 F

SN % , IleCl- -.01

INTERNATIONAL DEVELOPMENT RESEARCH CENTRE Ottawa Cairo Dakar Johannesburg Montevideo Nairobi New Delhi 0 Singapore

141 V /IL s 0

/'A- 0 . Preface

The International Workshop on Digital Radio Technology and Applications was a milestone event. For the first time, it brought together many of those using low-cost radio systems for development and humanitarian-based computer communications in Africa and Asia, in both terrestrial and satellite environments. Ten years ago the prospect of seeing all these people in one place to share their experiences was only a far-off dream. At that time no one really had a clue whether there would be interest, funding and expertise available to exploit these technologies for relief and development applications.

VITA and IDRC are pleased to have been involved in various capacities in these efforts right from the beginning. As mentioned in VITA's welcome at the Workshop, we can all be proud to have participated in a pioneering effort to bring the benefits of modern information and communications technology to those that most need and deserve it.

But now the Workshop is history. We hope that the next ten years will take these technologies beyond the realm of experimentation and demonstration into the mainstream of development strategies and programs. To the extent that integration occurs, we believe we will see these technologies used to reliably network those traditionally left out of conventional telecommunications services (the "last mile" problem) into the burgeoning and incredible array of information services now becoming available.

Through their efforts and vision, the readers of this volume will make real their response to the "last mile" (including perhaps even a future Workshop)! Acknowledgments

This workshop and its proceedings are a result of a significant grant from the International Development Research Centre in Canada, long-term planning collaboration and logistical support by Volunteers in Technical Assisstance, from both their headquarters in the United States and their Kenyan national office, an additional grant from the United Nations University and the intellectual contribution of all the participants.

Organization of this workshop was carried out in consultation with the organizers of the Workshop on Science and Technology Communication Networks in Africa - the African Academy of Sciences and the American Association for the Advancement of Science. In so doing, through holding the two workshops back-to-back in Nairobi, the degree of participation in both was enhanced. Contents

...... Workshop Background 1 - David Balson

Whither Digital Radio* ...... 3 - Gary Garriott

Development Applications of Packet Radio Technology: The New Challenge ...... 14 - S. Ramani

Digital Radio Applications in Terrestrial and Satellite Environments ...... 25

HealthNet in Zambia: The Technical Implementation of a Communications System for Health Workers* ...... 27 - Mark Bennett

Eastern and Southern Africa Network (ESANET), Communications Research for Development - The Experience of the Nairobi Node* ...... 41 - Anthony J. Rodrigues and Geoffrey M. Macharia

Digital Radio Technology and it's Application in Tanzania ...... 50 - William Sangiwa

Experiences in Setting Up a PACSAT Station in Tanzania ...... 54 - H.R. Mgombelo and S.J. Braithwaite

Sdminaire Radio par paquets, Note de Prdsentation, VITA/PEP ...... 62 - Yondailaou gue Tolloum

Mindanao Packet-Radio Network for Cooperative and Agribusiness Development ...... 71 - Josefito A. Guillermo

Packet and Satellite Radio in Sierra Leone ...... 90 - Joseph Sandi

How a Packet Radio Communications Systems has served ISERST* ...... 96 - Djama Dirieh

Packet Radio in Mozambique: One Failure One Success ...... 107 - Phil Gray Satellite-Based Developments ...... 119 University of Surrey/SST Presentation Notes ...... 121 - Martin Sweeting

VITA Operations using UOSAT-3 ...... 135 - Eric Rosenberg and Gary Garriott

SatelLife and HealthNet: The System and Its Operation ...... 145 - Dieter Klein

Regulatory Issues ...... 157 Kenyan PACSAT Licensing Experience ...... 159 - A.J. Rodrigues and G.M. Macharia

Summary of Discussion ...... 175

Training Issues ...... 177 Summary of Discussion ...... 179

Conclusions ...... 181

APPENDICES ...... 185

Agenda ...... 187

Participant List ...... 190

Additional Papers ...... 201 The United Nations University ...... 203 - Ines Wesley-Tanaskovic

The Big Opportunity: The Case of Healthnet in Zambia ...... 213 - Regina Cammy Shakakata

CEntral AMerican NETwork ...... 227 - William C. Hast

Experience With Low Cost Digital Communications (or Networking in and around the University of Yaoundd) ...... 231 - William S. Shu Technical Background Papers ...... 241

Error Correcting Radio Systems: A Discussion of SITOR and Packet Radio ...... 245 - Norman J. Sternberg

PACTOR - Radioteletype with Memory ARQ and Data Compression ...... 290 - Hans-Peter Helfert and Ulrich Strate

CLOVER Fast Data on HF Radio ...... 300 - Bill Henry and Ray Petit

HF Radio Data Communication ...... 304 - Bill Henry and Ray Petit

'Also published in: African Academy of Sciences, American Association for the Advancement of Science. Electronic Networking in Africa : Advancing Science and Tchnology for Development ; Workshop on Science and Technology Communication Networks in Africa, Aug. 27-29, 1992, Nairobi, Kenya. Washington, DC : AAAS, 1992. A workshop organized in consultation with the organizers of the International Workshop on Digital Radio Technology and Applications

Workshop Background

Through support for research, Canada's International Development Research Centre (IDRC) assists scientists in developing countries to identify long-term, practical solutions to pressing development problems. Support is given directly to scientists working in universities, private enterprise, government and non-profit organizations. IDRC's Information Sciences and Systems Division (ISSD), in the process of looking at telecommunications issues, became interested in packet radio technology over a decade ago.

Dr. Ramani (Tata Institute of Fundamental Research, India), who is a participant in this workshop, proposed the concept of a low-earth-orbit communications satellite to be built by developing countries for development communications at an ISSD-sponsored Workshop on Computer-based Conferencing in 1981. This workshop gave birth to ISSD's Telematics Program. This Program has promoted and supported the development, testing and evaluation of networking techniques and technologies in developing countries in support of development and research objectives. Since this 1981 workshop the Program has promoted this technology as a response to "last mile" communications problems.

The concept was explored further by Dr. Ramani and others: Richard Miller and Dr. Ramani co- authored a paper on the concept for ICCC '82. Ramani with the Indian Space Research Organization and Space Applications Centre; and with Yash Pal, Secretary-General of UNISPACE '82, at that Conference where one of its recommendations urged further developments in this field.

In late 1982, Volunteers in Technical Assistance and the Radio Amateur Satellite Corporation (AMSAT) picked up the concept and developed the Data Communications Experiment (DCE) payload as a prototype of their proposed low-earth-orbit electronic messaging satellite, PACSAT. The DCE was launched in 1984 on UOSAT-2 (a satellite built by the University of Surrey). An upgraded PACSAT Communications Experiment (PCE) was launched in 1990 on UOSAT-3. IDRC funded a demonstration of the DCE at the Pacific Telecommunications Conference in Honolulu in 1985.

In 1987, the Telematics Program funded the Lesotho Telecommunications Corporation (Lesotho's PTT) to carry out research on the viability of using ground-based packet radio technology for the provision of communication services in support of development applications in rural areas. VITA played an important role in assisting in the technology transfer process within this project. Unfortunately, due to a variety of problems, this project did not achieve all of its objectives, although it did prove the technical viability of the approach.

In 1988, a representative of the International Physicians for the Prevention of Nuclear War discussed communications and development with the Telematics Program. The PACSAT concept was suggested to IPPNW as a worthwhile concept to pursue.

IPPNW initiated a non-profit corporation called SatelLife to pursue this concept. SatelLife focused their efforts on three fronts: raising funds for acquiring communications capacity on the University of Surrey-built UOSAT-5; developing appropriate health-related information applications for the satellite (Healthnet); and beginning the process of identifying appropriate developing-country participants and of acquiring the necessary regulatory approval. In mid-1990 Satell-ife's Executive Director visited the participating institutions and individuals in the IDRC-supported ESANET project (this project involves networking research activities among five of the largest universities in Eastern and Southern Africa). Two more trips were made to

1 this region to discuss collaboration with the Telematics Program, the ESANET project, and IDRC's Health Sciences Division.

The above-mentioned discussions stimulated further consideration of project activities among several IDRC Divisions. As a result, IDRC acquired access to communications capacity on UOSAT-5 and approved funding for a research project entitled "Healthnet: Satellite Communications Research for Development" with the following objectives: to empower researchers in the health field (focusing on Africa) through the provision of access to information and communication capabilities; to build research support capacity in health-related developing-country institutions; and to test, demonstrate and evaluate the use of packet radio satellite communication technologies in support of health information flows and networking in order to determine the sustainability and appropriateness of this technology. Details concerning this project will be presented during this workshop.

Since 1986 VITA and IDRC have been discussing the need and appropriateness of organizing an international workshop on the subject of digital radio technology and its applications. Indeed, IDRC approved funding for the workshop in early 1989. The original plan was to hold the workshop in Tunis, Tunisia in collaboration with the Centre National de I'Informatique. Efforts were made to hold it in concert with other larger conferences to ensure greater participation. Finally a date in early 1991 was established; travel problems which arose as a result of the Gulf war caused a further postponement and a rescheduling to August 1992 in Nairobi, Kenya.

IDRC has been involved with the promotion of the utilization of this technology for a long time. The interest and excitement has not been solely related to the technology itself, but rather on its cost-effective use and application for development. Related projects present an opportunity for testing the use of communication and information technology as tools for empowering developing-country institutions and individuals. It provides the opportunity to demonstrate to telecommunication authorities.a cost-effective, interim (10-20 years) technology which could solve some of the current infrastructure problems experienced in developing countries, especially in Africa.

With these words, after many years of effort on the part of many persons, 1 sincerely welcome you to this workshop and have every confidence that we will all find this an informative and worthwhile event.

2 WHITHER DIGITAL RADIO?

Gary Garriott Director Informatics

Volunteers in Technical Assistance 1600 Wilson Boulevard Suite 500 Arlington, Virginia 22209 USA

Prepared for:

International Workshop on Digital Radio Technology and Applications

Nairobi, Kenya

August 24-26, 1992

3 Whither Digital Radio?

An answer to a technical problem that takes minutes to obtain in Europe can take months to obtain in Somalia or Sudan. To give iust one example, a medical advisor in Mogadishu needed background information on excretion of antimalarials in breast milk to help him decide on the details of a prophylaxis programme for about half a million people. The agency funding him had no staff in Europe who were themselves qualified to make a thorough search for this ifnormation or who knew who to ask to do it for them. The telephone calls necessary to-set up and pay for a search through a Western information centre would have taken weeks, given the communications problems at that time. The solution was to get a friend who was. passing through Nairobi to pay himself for a search in Europe, personally photocopy the papers concerned, and then to mail the printout and copies of papers to Mogadishu. The total time needed to get the information on this routine enquiry was about four weeks. The programme was already underway when the material arrived. Hundreds of highly technical decisions affecting huge numbers of people are made every month in relief prorgammes with a bare minimum of scientific background data.

So writes R.S. Stephenson in the October 1986 issue of

Disasters. His is a graphic way of stating that the accuracy of information is an important but insufficient condition for its use, especially in developing countries. In order for most technical and

logistical information to be employed in the execution of a

project (or in this case, a relief operation) it must be timely as

well as accurate in content. Scientists, engineers and physicians--

frequently the source of crucial knowledge--are usually aware of

the time dimension of technical information requirements in their

own jobs; however, they may may not always appreciate that good

timing is also an operative factor in relief and development

projects.

The lack of reliable communication with remote

4 simply, the packet controller is the distinguishing device which marries the personal computer to the two-way radio. The TNC performs many functions including error- checking, where short bursts of audio tones containing the digital information are checked. Unlike normal fax and some other digital transmission modes that do not check for and correct errors, the high reliability with which packet radio encodes and decodes digital information means that sophisticated computer programs, data files and even graphics can be transferred from one point to another accurately.

When using packet radio, the radio connections will be about as good as they are on voice radio, that is, packet transmissions will probably be acceptable if voice radio contacts are strong and intelligible. If signals are weak or interference and static are high, packet radio will not work as well.

Additionally, some radio frequencies are more suited for packet transmissions than others. These are the VHF

(very high frequency) which range from approximately 30-

300 megahertz (MHz) and UHF (ultra high frequency) which range from 300-3000 MHz. While "high frequencies" (HF), typically 3-30 MHz, are less ideal, distances between stations can be much longer (1-3,000 kilometers) as compared to VHF/UHF which are typically within 100-150

5 regions has posed a difficult obstacle in the implementation of development projects for decades. While regions in Asia, Africa and Latin America are expanding communications channels through modern digital switching equipment and even fiber optic technology, many rural areas continue to be isolated. The development of telephone circuits accessible to economically marginal groups is occurring slowly, when at all. Inexpensive digital technologies such as packet radio, however, while not replacing the lure and utility of voice telephone, may now be a viable low-cost alternative in bridging the

"last mile" gap which plagues the distant client or the 4 end-user with specialized requirements.

What is Packet Radio?

Packet radio combines two mature and relatively low-

cost technologies -- two-way radio and personal computers

-- in a system that permits the computers to communicate with each other over radio circuits. Analogous to

computer communication via modem over telephone lines,

packet radio is easy to use and permits the transmission

of messages, letters, spreadsheets and reports without

the need for manual transcription or intervention. Packet

radio networks can be simple or complex, depending on

communication needs.

The "terminal node controller" (TNC) or, more

6 kilometers.

In theory, it is possible to adapt existing two-way radio stations to packet radio stations when certain characteristics are present. In most cases, new radios are preferable because later technology more easily accomodates packet communications. An analysis of the current system can determine the feasibility of using existing technology.

What Has It Done?

Military uses of packet radio have been widespread.

Persistent rumors have claimed that users of packet radio technology have included RENAMO in Mozambique, the

Eritreans in Ethiopia prior to the recent overthrow of the Mengistu regime, the contras in Nicaragua, and some marxist factions in the Philippines. Anti-drug packet radio networks have also been cited in operation throughout Latin America.

The first known humanitarian use of packet radio was by VITA in 1986 when two VITA volunteers traveled to

Ethiopia invited by CARE and the Ethiopian Relief and

Rehabilitation Commission. A demonstration between CARE

offices in Addis Ababa and Dire Dawa, several hundred kilometers to the north, successfully exchanged

logistical information on food supplies for several weeks. Since then VITA has installed such "terrestrial" networks and trained local staff in the Sudan, Madagascar,

Philippines, Chad and Jamaica. Additional studies and demonstrations have taken place in Nigeria, Tanzania,

Mozambique, Pakistan, Afghanistan, Lesotho, and Kenya.

The packet networks in Jamaica and the Philippines have extended preparedness and search and rescue information to isolated areas during and prior to disaster relief efforts.

What Else Can It Do?

Non-commercial packet experimentation in Latin

America and in India has also been extensive. Most applications involve the transmission of administrative messages which either required frequent repeats or the use of multiple languages and have, therefore, been

inefficiently transmitted by voice radio, if at all. In

these cases, packet radio facilitates the provision of a

hard copy. Because non-text files such as spreadsheets

and database result sets can also be transmitted, these

administrative applications are likely to increase as

computer skills beyond wordprocessing continue to expand.

Perhaps the most exciting application of packet

radio technology is the adaptation to store-and-forward

messaging via inexpensive satellites in low-earth orbit.

When in polar orbits, such satellites traverse all points

8 on earth at least twice a day. During these "passes" -- typically ten to fifteen minutes long --hundreds of pages of text (or the equivalent in other kinds of files) can be uploaded or downloaded to addressees somewhere else in the world using currently available technology. VITA has been a pioneer in this application, beginning in 1983 when the "Digital Communications Experiment" developed by

VITA staff and volunteers was launched on a University of

Surrey (UK) spacecraft. In 1990 the "PACSAT

Communications Experiment" was launched, sponsored by VITA to operate on special frequencies. Today development and relief-related demonstration groundstations are being installed by VITA and others, such as SatelLife, throughout the world, with the emphasis on Africa. Space applications of packet radio currently emphasize information exchange and dissemination on topics in health, education and energy/ environment as well as administrative and logistical information related to relief and development programs and projects. The launch of VITASAT-A in.the early 190's will mark the first low earth orbiting communications satellite in history totally dedicated to humanitarian ends. In

1992, VITA received the first ever "Pioneer's Preference" award from the Federal Communications Commission for its work. VITA was very active in negotiations at the World Administrative Radio

Conference which for the first time allocated VHF and UHF spectrum to low-earth orbiting communications satellites.

9 What About Installation and Costs?

When choosing between terrestrial and space applications of packet radio, generally, within-country communication is best achieved by ground-based packet radio networks, while packet radio in space is usually preferred in between-country communications. Unusually large countries (Sudan) or dispersed nations (Kiribati) may also benefit significantly from packet radio satellites.

Whether in space or terrestrial environments, experience thus far indicates that a minimum of one week per station is necessary for installation and training.

Two levels of training must be provided: operator and

"systems engineer." Operator training is usually accomplished within a matter of hours with anyone who has some DOS and wordprocessing computer experience. "Systems engineer" involves trouble-shooting and problem-solving.

This experience is generally provided over a period of some weeks through "on-the-job" practice and tasks assigned by the more experienced installers. A moderately-skilled radio or computer technician is a prime candidate for the system engineer level of responsibility.

Determining the specific variables and cost needed

10 to install and operate packet radio requires a study of the environment. Excluding the computer, terrestrial packet radio station hardware and software can cost anywhere from $2000 to $10,000 each (installed), but the cost depends largely on the quality of the radio required. For example, a cheaper radio could be used in a desert environment where interference is low.

Comparable "fixed" and "portable" satellite stations can range from $1,500 to $5,000. In both terrestrial and space environments, important variables include the quality of the antennae system, and the availability of stable electricity from mains or generators or from solar panel/battery combinations,

Commercial TNC and radio manufacturers, primarily in

North America, Europe and Japan, are hotly competing for government and military contracts overseas, where per station costs can easily exceed $20,000. To our knowledge, VITA is the only organization that has been promoting lower cost alternatives by adapting hardware and software made available through the amateur radio marketplace for relief and development applications. VITA is able to provide a series of packet radio-related services, from conducting needs analyses and feasibility studies to actual installation and training and post- project technical support.

11 What is the Future of Packet Radio?

The greater the links, the greater will be the utility of packet radio. Not only can terrestrial networks separated by great distances be linked through packet radio satellites, but they can also be interconnected with inexpensive "landline" (telephone) networks, such as FidoNet and BITNET or the Internet.

Constant advances in technology indicate a continued dynamic and growing marketplace. Unfortunately, security concerns and regulatory issues cloud the rapid dissemination of this technology in many developing nations, even for humanitarian ends.

Packet radio is not a panacea for solving the age- old communication dilemmas from remote areas. It presupposes that computers have been introduced for reasons other than communication and that indigenous skills have reached levels where it is natural to transfer computer-generated files and programs to other locations. It also assumes that a progressive-minded government is investing in and providing its citizenry with more of the potential benefits of information technology, including those at the socioeconomic periphery. Much still needs to be learned about how these systems work best given disparate geographical and socioeconomic settings.

12 The use of packet radio technology is one way to ameliorate the marginalization associated with living in distant regions by providing connections to other individuals and networks, nationally and internationally.

Packet radio is a "hot" personal medium, meaning that

"real human callers" (popular electronic mail terminology) are present on each end. For this reason, the technology deserves the attention and scrutiny of those committed,to the expansion of information delivery systems toward the evolution of person-mediated

"knowledge networks".

Dr. Gary L. Garriott Director Informatics VITA 1815 N. Lynn Street Suite 200 Arlington, VA 22219 (703) 276-1800

The opinions expressed are solely those of the author and do not necessarily reflect those of Volunteers in Technical Assistance.

*Other digital radio technologies, such as Baudot, TOR (Teleprinting Over Radio), and ASCII also exist, with TOR being perhaps the more useful. While TOR does not support file exchange and is very slow, it is less sensitive to "noisy conditions" than packet and is widely used. Other papers consider TOR. VITA's experience to date has been almost exclusively with packet radio.

13 Developmental Applications of Packet Radio Technology The New Challenge *

S. Ramani

National Centre for Software Technology Guhnohar Cross Road No. 9 Juhu, Bombay - 400 049, India Email: ramani0saathi.ncst.emet.in

Abstract

This paper covers two topics: a) technology profitable for use in data communication for development catalysing applications, and b) the specifics of a few such applications.

The thesis advocated is that it is now time to turn attention away from purely technical issues to real applications. The cost-effectiveness of technology used cannot be ignored. Substantial progress has been made around the world in making data communication facilities widely available. Special arrangements cart reduce costs, but that is not the most important concern in this field. Significant applications which can improve the life of a significant number of people are the main concern. We do not have an adequate number of such applirattolls in use. This p alorr also focuses attention ore mass applications, something that has bccrt often ryrtmrd. There are hundreds of millions of people around the world, whose lives will be better if attention is paid to giving them specially designed services based on data communication. But very often, specialised datacom based services are planned only to benefit small groups of people. This 1xtpcr also deals with communication services for people in remote areas as an end objective. Non-tradatonal communication systems have to be invented to meet the needs of large numbers of people. Those concerned with development of remote areas, rural areas and regions facing special problems should plan to innovate and create new forms of communication systems.

'T'aper prepared for presentation at the International Workshop on Digital Radio Technology and Applications. August 24-26, 1992, Nairobi, Kenya.

14 1 Background

What is often taken for granted in a modern urban setting does not exist in a rural or remote area. In our context, we should consider the following:

reliable communication over short distances - tens of kilometres

reliable communication over long distances - to distant cities

affordable communication

widely available communication

appropriate communication - for instance, faxing of messages in a local script

Amazing developments in communication have changed the modern urban environment by providing:

reliable and affordable telephone networks

inexpensive data communication through dial-up links and through data networks

integration of office word-processing and record keeping facilities with networks

Nothing even remotely like this has happened in the rural. and remote areas. Even the urban areas of developing countries do not have facilities like this. Many of us have, as a result, worked on alternatives for developing countries. This has included efforts to find alternative satellite communication options [Ramani and Miller, 1982], and use of message systems [Gupta and Ramani, 1981. It is necessary to review our efforts periodically, and to redefine our goals. This paper is an atteirll)t to do this in the context of the technology of 1992. This paper is a second step in the direction indicated in a companion paper [Ramani and Gupta, 1991].

2 Improved Reach of Computer Networks

A number of developing countries already have access to a number of computer networks. A table (Landweber, 19921 provides information on network connectivity round the world. An abstract from this table is given in the Annexure. The main question we have to ask ourselves in this context is `Do we need specialised networks, or can we use a standard network for this particular application?' There is a great deal to be gained by development catalysing applications, such as connecting a teaching hospital or an agricultural research station to a network. Wherever this can be achieved by connecting the institution concerned to the world-wide academic 15 network, we should prefer to do that. The world-wide academic network is partly the Internet. It is partly the X.400 network used by the academics. Both these are, anyway, interconnected. The wealth of information available on the academic networks should not be un- derestimated. Access to library resources, access to distributed bulletin boards, to world-wide mailing lists, etc, are available now. An equally important fact is that over 5 million users in over a 100 countries are accessible over this network. Another equally important fact is that the Internet is connected to the R&D laboratories of a large number of commercial organisations, publishers and book sellers. Further, a commercial Internet exchange is being set up, enabling commercial applications of the Internet.

3 Satellite Communication

The concept of a geostationary satellite communicating directly to one's home is no longer a dream, it a reality. What are possible datacom applications of this technology ? What does it offer for development communication ? Can we pipe electronic bulletin board information to thousands of libraries using this technology ? Can we broadcast information on books being published to these libraries on a daily basis ? Can we send them current contents of periodicals ? Can we share abstracts of papers being published ? Can we even send text of journals to those who pay a subscription, perhaps using some form of encryption to enforce necessary distribution control ? Can we distribute the hundreds of electronic journals already in existence using this technology ? Can we distribute public domain software ? Can we distribute courseware ? Can we combine voice with data, and share seminars, conferences, and courses with the world ? Can we use specialised video techniques to provide limited video capabilities? In short, can we have a lamp in the sky which will illuminate us with the shared knowledge of humanity ? On-line ? Obviously, yes. Perhaps the UNESCO is the right body to consider such a scheme, to tie up thousands of libraries in the world, particularly in those countries where satellite communication is a necessity, into a knowledge sharing network. We could perhaps call the network the 'sun', to stand for 'super university network'! The task is doable. There is no major legal problem.that cannot be solved. For instance, there is no unsolvable problem in respecting the intellectual property rights to the information being transmitted. This problem is being routinely solved by everyone who operates a network on a regular basis. Those who have to be paid for their contributions can be paid, and those who are willing to pay for it can be persuaded to do so. The system does not have to make everything transmitted free for all. At the same time, we do have to recognise that humanity still gives a lot away. Scientific publications constantly give away a wealth of knowledge. Universities do

16 not always want to make money out of their contributions to human knowledge. Government labs often have valuable things to give away for reasons of their own. Let us note that most of what is being mentioned here is being done constantly by computer networks. But what is missing is that those who are behind are constantly falling behind. Those who are not on networks are being forgotten. Those who need knowledge most get the least. So, it seems logical that we ought to use the same technology that sells soap around the world today to share knowledge. We can follow satellite television stations very closely. We can lease a transponder and cover a big continent. We will need to run a packet broadcast network using the satellite, with very small apperture terminals feeding packet broadcast receivers at thousands of locations. We can use three thousand dollar systems to give every library access to a new system of communication. We can lease a small number of transponders to create a truly global system, not merely for the grandeur, but for real use and the symbolism. To tell everyone that, at last, the world of knowledge has become a single network. That we share precisely the same stream of knowledge, which we all work to create. This network need not be purely for library users. Obviously, it can carry health information to hospitals, agricultural information to the field stations and so on. All this may not even need a full transponder for each region. A fraction of a transponder might be enough.

4 The Dial-up Alternative

The telephone network is an economic necessity. So, more and more corners of the earth arc acquiring good quality telephone systems. Electronic exchanges and satellite earth stations quickly bring in new nodes into the world telephone network. Dramatic improvements have been seen in countries like India, with hundreds of electronic telephone exchanges being commissioned every year. In parallel, the technology of dial-up modems is making rapid strides. For instance, V32bis modems with V42 and V42 bis standards for error handling and compression arc revolutionising dial-up data communication. Inexpensive software makes dial-up Email very practical over hundreds of miles. Rapid strides in modem technology arc said to be partially responsible for slowing down the move to Integrated Services Digital Networks (ISDN). You do not need to integrate because integration is a good word ! You integrate when it is valuable. If you can get by using good modems, you can postpone expenditure on integration ! I would argue that simple technology like dial-up should not be ignored. Simple technology coexists very well with sophisticated technology, particularly in computer communication. In fact, for over a year we ran a courier service twice a week to a prestigious educational institute in India. The Institute had thousands of students and hundreds of faculty working in different areas of modern science and technology. But they were relatively isolated, being well outside a city not well connected to the 17 nation's telecom network. Hundreds of Email items were carried on floppies and read in. They were distributed over the campus by a sophisticated local area network, and delivered through dozens of computers to their users. Network users all over the world knew that they could now contact their colleagues at this campus, but did always not know that a floppy-by-courier system was doing the real work ! It did not matter, for the old airmail delay of four weeks for a letter and answer was replaced by an average round-trip delay of four or five days. Email came to Bombay by computer network, and was passed on by floppy-by-courier. Replies came back to Bombay the same way and then went on to the fast track of computer network again !

So, the lowly dial-up modem need not be forgotten. It can coexist with very so- phisticatcd systems. It can often provide the only link. Sometimes, it will provide an acceptable end link. The situation in regard to dial-up data links has improved very significantly in many places in the world.

5 The Challenge of Mass Communication

Improving communication is a true developmental objective. We should not accord communication a priority lower than health or education. Communication can play a catalytic role in improving the economy and quality of life in a rural area. There seems to be, unfortunately, no alternative as yet for meeting the communication needs of the rural population. Telephony continues to be far too expensive for this purpose in developing countries. Packet radio can offer us this alternative, provided it is combined with appropriate delivery mechanisms such as Fax. One of the reasons why Fax is so attractive is that it allows people to communicate in their own script. Further it allows painless sharing of a communication link, unlike a public telephone, by a number of people. For instance, consider a combination of

a V11F transceiver

a microprocessor handling a packet radio protocol a Fax-like interface for handling written communication

The whole thing could be battery powered, and vehicle mounted. Such a system can easily communicate to a district centre 30 km away, serve dozens of villages and provide fast and inexpensive com munication. At the city end, the system would be connected up with the nation's postal system, as well as the nation's Fax network. This would be appropriate technology in a true sense. It would be very practical as a low-cost system and can be started up without major expenditures. It would also enable mass production of the equipment, as large numbers would be required to meet the needs round the world. 18 6 Conclusion

The focus in this paper has been on mass applications of packet radio technology for meeting development objectives. It has been argued that exploitation of the rapidly spreading computer networks round the world has not been satisfactorily exploited for this purpose. The purpose being more important than the technology, where we can get by without packet radio technology, we should do so. There have been tremendous improvements in the dial-up data communications. As persons interested in packet radio technology, we can take pleasure in noting that this has been the result of using packet switching technology over the telephone network. What we do over the radio, or the satellite link, they do over the telephone network. Error correcting modems and high-speed modems have conquered the telephone network. We can again exploit this technology whereever possible. The value of using widely available channels made available by geostationary satellites has been pointed out. The possibility of a revolution comparable in magnitude to that of the satellite television revolution has been discussed. Lastly, it has been argued that communication is a significant development objective. The possibility of using packet radio technology for mass applications in rural communication has been discussed.

References

[Gupta and Ramani, 19811 Gupta PP and Ramani S. Computer j1lcssage Systems for Developing Nations: A Design Exercise. Prescntcd at the International Symposium on Computer Message Systems, Ottawa, Canada, April 6-8, 1981. Proceedings Ed. R. Uhlig, North Holland, 1981. [Landwcbcr, 19921 Landwcbcr L. International Connectivity. In Internet Society News, Spring 1992, Vol 1, No 2, pp 99-52. [Ramani and Miller, 19821 Ramani S and Miller R. A Neu, Type of Communication Satellite needed for Computer Based Alessaging. Presented at the Sixth International Conference on Computer Communications, London, Scl)tcmbcr 1982. Proceedings: North Holland, 1982. [Ramani and Gupta, 19911 Ramani S. and Gupta P.P. Computer Communication for Development Some Challenges Ahead. Presented at the Conference on Computer Communications for Developing Coun-

19 tries, (CCDC-91 Tunis, May 21-23, 1991.) Proceedings: North Holland, 1991.

20 Annexure

INTERNATIONAL CONNECTIVITY

Version 1 - August 8, 1991

Please send corrections, information and/or comments to:

Larry Landweber Computer Sciences Dept. University of Wisconsin - Madison 1210 W. Dayton St. Madison, WI 53706 lhl@cs.wisc.edu FAX 1-608-263-7442

Include details, e.g., on connections, sites, contacts, protocols, etc.

Thanks to Rick Adams, Randy Bush, Susan Calcari, Donnalyn Frey, Michael Gettes. Frode Greisen, Daniel Karrenberg, Enzo Puliatti. Florencio Utreras, Dave Wilson for their input.

In the following. BITNET is used generically to refer to BITNET plus similar networks around the world (e.g., EARN, NETNORTH. GULFNET. etc.).

------The following list has been edited to exclude countries which are not developing countries. But, no precise classification has been attempted.

Ramani ------

SUMMARY

NUMBER OF ENTITIES WITH INTERNATIONAL NETWORK CONNECTIVITY = 84

BITNET (Entities with international BITNET links.) = 1 domestic site 5 > 1 domestic site = 41 INTERNET (Entities with international IP links.) operational = 31 uncertain = 1 21 UUCP (Entities with international UUCP links.) minimal < 5 domestic sites = 41 widespread >= 5 domestic sites = 37 FIDONET (INCOMPLETE) (Entities with international FIDONET links.) minimal < 5 domestic sites = 34 widespread >= 5 domestic sites = 2 Col 1. BITNET INTERNATIONAL LINK - = no link b = 1 domestic site B >= 1 domestic sites Col 2. INTERNET INTERNATIONAL LINK - = no link = uncertain I = yes link Col 3 UUCP INTERNATIONAL LINK - = no link u = minimal < 5 domestic sites U = widespread >- 5 domestic sites Col 4 FIDONET INETERNATIONAL LINK - = no link f = minimal < 5 domestic sites F = widespread >= 5 domestic sites

BIU- AR Argentina (Argentine Republic) --u- BO Bolivia (Republic of Bolivia) ---f BW Botswana (Republic of Botswana) BIUf BR Brazil (Federative Republic of Brazil) --uf BG Bulgaria (People's Republic of Bulgaria) --u- BF Burkina Faso (formerly Upper Volta) --u- CM Cameroon (Republic of Cameroon) BIU- CL Chile (Republic of Chile) --u- CI Co-te d'Ivoire (formerly Ivory Coast) B-U- CO Colombia (Republic of Colombia) b-u- CR Costa Rica (Republic of Costa Rica) --u- CU Cuba (Republic of Cuba) b-U- CY Cyprus (Republic of Cyprus) --u- DO Dominican Republic B-u- EC Ecuador (Republic of Ecuador) B-u- EG Egypt (Arab Republic of Egypt) ---f ET Ethiopia --u- GF French Guiana (Department of Guiana) --u- PF French Polynesia --u- GP Guadeloupe (Department of Guadeloupe) --u- GT Guatemala (Republic of Guatemala) BiU- IN India (Republic of India) --u- ID Indonesia (Republic of Indonesia) B-u- MY Malaysia 22 --u- ML Mali (Republic of Mali) --u- MQ Martinique (Department of Martinique) --u- NA Namibia --u- NI Nicaragua (Republic of Nicaragua) --u- NE Niger (Republic of the Niger) --u- PY Paraguay (Republic of Paraguay) --u- PE Peru (Republic of Peru) --u- PH Philippines (Republic of the Philippines) --u- zz Polynesia (French Polynesia) --u- RE Re'union (Department of Re'union) --u- RO Romania (Socialist Republic of Romania) --u- SN Senegal (Republic of Senegal) --u- SC Seychelles (Republic of Seychelles) -- - SL Sierra Leone (Republic of Sierra Leone) --u- LK Sri Lanka (Democratic Socialist Republic of Sri Lanka, formerly Ceylon) --u- TG Togo (Togolese Republic) bIu- 'TN Tunisia b-u- UY Uruguay (Eastern Repuplic of Uruguay) --u- VE Venezuela (Republic of Venezuala) ---f ZW Zimbabwe (Republic of Zimbabwe)

23 24 Digital Radio Applications in Terrestrial and Satellite Environments

25 26 HealthNet in Zambia: The Technical Implementation of a Communications System for Health Workers

Mark Bennett Computer Centre University of Zambia

1. Introduction: The Objectives of HealthNet

In 1985 Dr Bernard Lown (co-president of the Nobel) Peace Prize winning organisation IPPNW) conceived of the idea of an organisation that would use micro-satellite 'Star Wars' .technology for peaceful purposes: the sharing of health information amongst professionals, particularly in developing countries. In 1989 this conception became reality with the founding of 'SatelLife', a joint east-west organisation (based out of the USA and the then USSR) dedicated to building north-south and south-south medical communication.

The objective of the project was strong and cohesive. It was known that 'information' was one of the commodities of which developing countries were most short: information itself being an expensive commodity with a large community of 'have-nots' in a similar disparity to economic inequality. Lack of access to health information leads to loss of life: it leads to replicated research: it leads to lack of awareness of simple but innovative cures or preventative meas- ures. This is especially so in SatelLife's initially targeted area of Africa where basic health infrastructure may be weak or underfunded and where the disease AIDS is assuming severe proportions.

It was also known that traditional forms of communication were not always effective. The telecommunications infrastructure within African countries varies greatly. Through- out the region as a whole (from country to country) it is weak and expensive. To take Zambia as an example, an international telephone call to the UK costs in the order.of USS7 per minute and calls to neighbouring countries (where they are possible at all) will cost up to half of that. This is a high cost in comparison to a few cents for a packet of Paracetamol or a few hundred dollars as the monthly salary for a doctor. In addition it must be remembered that telephone density is low in Africa (in Zambia it is around 0.7 phones per 100 people: less that one hundredth of the North American figure). Conventional mail is very slow (up to 3 weeks to -a remote part of the country); transport may be poor and expensive. Libraries may have stopped taking recent medical journals through lack of funds or foreign exchange.

It was some of these considerations that caused SatelLife to adopt packet radio technology combined with a low-earth orbiting satellite which could act as a 'postbox' to store

27 and forward mail as well as distributing appropriate medical information which was felt to.be relevant. This technology was felt to provide a relatively low-cost solution to the communications problem which would by pass national telecomms infrastructure problems and allow communication on an equal basis to and between all countries interested in joining the project.

Zambia was fortunate to be one of the first countries within the sub-region to install the equipment provided by Satel- Life. This paper addresses some of the technically related aspects of Zambia's experience in establishing its Health- Net.

2. The Basic Technology.

The satellite which is used by the HealthNet project is similar to that used by the amateur radio community. It 'is low-earth orbiting (in an orbit some 850 km from the earth) which means that only a small low-powered transceiver and antenna are required, in distinction to geostationary satellites. The cost of such satellites is small in comparison - some USSlm to build and launch (perhaps 1/400th of the cost of a telecommunications satellite). The HealthNet satellite was developed by the University of Surrey in the UK. Being 'sun-synchronous' it orbits at a speed of some 17,000 kph and passes over each point on the earth at least 4 times a day for an average of about 10 minutes per pass.

The satellite, weighing some 40kg has about LOMB of on board memory, designed to store messages for around 2 days. In theory it should of course be possible that a message send by any one earth station will be picked up at the destination within half a day. One pass should be sufficient to pick up perhaps 100k of data, although in practice this has probably not been achieved by many stations to date.

The satellite transmits a continuous signal denoting its presence. This is then picked up by the receiving site which responds, and handshaking starts which instigates a communications session. The basic equipment required on the ground is a PC (with hard disk; 12 MHz or faster, preferably with a maths coprocessor), an amateur radio transceiver, and a terminal node controller with integrated radio modem. In addition separate transmit and receive antennae are used: currently of the egg-beater design. The cost of the whole groundstation is thus less than USS5,000 including the computer.

In some locations electricity supply may be a problem, but in theory it is possible to run the whole operation from solar power.

28 It will be appreciated that, since the groundstation is communicating directly with a satellite, and particularly since it is transmitting rather than just receiving, that permission has to be gained from the local licensing authority (normally the PTT) to operate the system within any country. Whilst worldwide authority will have been sought by SatelLife for use of the relevant frequencies these must also be cleared within individual countries. In addition certain countries may have concerns about whether the PTT itself is being by-passed or whether there are any security implications.

3. Local Infrastructures

Fortunately Zambia obtained a licence to operate the station in a relatively short time scale, with generous assistance from the Zambian PTC and bearing in mind the humanitarian work that was to be done by the system. Despite initial fears licences have now been obtained in many other countries. Groundstations are now operational in Zambia, Kenya, Uganda, Tanzania, The Congo (where the African Regional Headquarters of the WHO is sited) and Mozambique. Zimbabwe is expected shortly and Botswana is also hoping to join the scheme.

It will be appreciated that although SatelLife have set up an infrastructure that is beginning to provide international communication where it did not previously exist, that within each individual country there will also be a need to share the information and allow a variety of widely scattered users to gain access to the communication channels newly opened. It is possible for there to be several groundstations within a country although this is an expensive solution and causes contention for the satellite time. Where telecomms facilities are very poor this is one option that will have to be considered, along with the use of 'line of sight' terrestrial based packet radio to link points to the main groundstation.

In Zambia the option chosen has been a Fidonet-based network using computer-to-computer links through normal dial-up public telephone lines. The majority of institutions likely to require eventual link-up with the system have a telephone that is normally operational for the greater part of the time, and available for the necessary few minutes a day to exchange mail (fido uses batch-based rather than interactive mailers which maximises the use of poor or expensive phone lines and requires the minimum human intervention). The sites also have an electricity supply for a sufficient time each day to operate the computer.

The satellite groundstation is located at the University of Zambia Computer Centre which also houses a dedicated email PC running Front-Door Fidonet software (used for the Univer-

29 sity's own email system). A standard modem within the groundstation computer allows satellite-based messages to be exchanged between these two systems. Presently there are two email stations operational in the University Teaching Hospi- tal (based in central Lusaka some 8km from the main campus); two stations at the Ministry of Health; one at Lusaka Of- fices of the World Health Organisation (WHO); one at the UNICEF Offices in Lusaka; one at the Tropical Diseases Research Centre (TDRC) in Ndola (some 300km away); and one in Nkana Mines Hospital in Kitwe (370km away). These stations keep up regular contact with the satellite and with each other.

The project plan (being coordinated by a HealthNet planning team consisting of representatives from the University, the Ministry of Health - who will ultimately assume responsibility for the system, WHO, UNICEF and CMAZ) is to begin with a pilot project in the Southern Province of Zambia, connecting some 9 District Hospitals to the network. These hospitals consist of a variety of sizes (Mission and State-run), but all could be considered to be serving the rural community. This pilot phase will test the viability of extending the system to all 84 hospitals in the country. This work is being done in conjunction with a project under the Ministry of Health which was to use unlinked computers in some provincial hospitals to collect management data.

The major tasks of the system to be implemented will be:

a) Improvement of local, regional and international communication with others in the medical and health profes- sions.

b) Use in health service administration, including planning, budgeting, accounting, manpower control, provision of health statistics and vehicle control).

c) The improvement of distance education, including in- service training to doctors.

d) The provision of up-to-date literature, both local and international.

e) Assistance with control of the drug supply to hospitals and clinics.

f) Communication regarding clinical consultations and referrals, as well as available treatment facilities and policies.

g) The gathering and subsequent distribution of epidemiological information which could also be made available to international organisations such as the WHO.

30 In connection with the literature facilities above, generous assistance has been received from the Health Foundation in Washington which has provided a 4 stack CD-ROM system, along with a subscription to Medline on CD to attach to'the PC in the University School of Medicine Library. This will in turn be able to 'feed' the network with some 'in country' literature requests. This has been complemented by AAAS who have supplied further CDs and by a link between the University Medical Library and a counterpart at the University of Flori- da who are able to supply (by email) articles not available locally. A database search 'online' through fido has yet to be established, but is feasible once suitable software has been installed.

4. Successes and Failures

It has been possible for the University of Zambia to set. up the groundstation and get it, and the supporting Fido et- work, operational without any staff having to come in from outside to set things up or keep them running. The equipment was donated from SatelLife and IDRC but no funding was available for staff, and it should be possible to repeat this success in any other suitable and comparable situation.

This is a relatively new technology being operated, in the case of SatelLife, by a relatively new organisation. It is likely also to be a transitory technology (the take-up on microsatellite technology appears to be growing fast and may emerge quickly into the commercial sector). There have thus inevitably been many learning experiences to go through and problems to encounter. Some of these are listed briefly below for the benefit of others who may be working in similar areas. They are divided into technical and human resource problems.

4.1 Technical Problems

a) The satellite passes overhead 4 times each day (more nearer the poles, but 4 is the figure for most of Africa). 2 passes are during the late morning and 2 just before midnight when the station will be unmanned. Testing and tuning the system is thus restricted to a maximum of 20 minutes a day and changes made to antenna positions etc. will not be possible to be checked until the following day. Some of this may be solved when a 'satellite emulator' board is produced as may eventually be the case.

b) The system is designed to provide communications for locations where communication is currently poor. There is thus something of a Catch-22 situation when the system does not function as it should: how do you communicate with someone who can provide help? Zambia was fortunate In that a terrestrial fido link with

31 first Greennet (in London) and then the Internet (through Rhodes University in South Africa) was established concurrent with HealthNet. The fido system was not without its problems (and does not provide communication with all the surrounding countries covered by HealthNet) but it did provide a one-day turnaround de- bugging link with the SatelLife staff in Boston. Debug- ging at a distance can still be hard, but much easier than by snail-mail!

C) Being a newly developed (and developing) combination of technologies the documentation associated with the system is also developing and has not yet reached the professional stage of some computer systems. This is particularly so with the software which has changed frequently. d) Several discoveries have been made with the running of the system including the need for extreme accuracy in certain areas. The Doppler effect- is important in calculating the frequencies with which the satellite signal will be received (and with which it needs to be transmitted). This has meant that the computer clock must be accurate to within one or two seconds of actual Greenwich Mean Time. Likewise the satellite characteristics, as defined in the Kepplerian elements, need updating at least once a month. It is possible that the time might be set directly from the satellite in the future, and also that the Kepplerian elements may be automatically downloaded. Both these steps would assist in reducing the manual intervention currently required. e) There are considerations regarding the optimum (or maximum) length of cable between the transceiver and the antennae. Once this is exceeded signal loss becomes unacceptable. The Zambian groundstation located its antennae on the top of the third floor roof (with the computer on the ground floor) and the signal loss on the receiver in particular was too great. A single drop low-loss cable has since been put in which has improved matters although full re-siting of the computer may be necessary.

f) The siting of the equipment, particularly the antennae may be critical. They must have an uninterrupted vision of the horizon in each direction wherever possible. Within the environment found in a number of countries the security of the equipment itself has to be considered, so the antennae again cannot be left exposed and unsecured. Being frail they might be effected by severe weather conditions in certain localities, and they are clearly susceptible to tropical lightening strikes when placed high on a building.

32 g) Whilst the PC which controls the running of the system may be serviceable locally it is unlikely that the same is true of the transceiver. Skills may or may not be available to maintain the TNC/modem or the antennae (noting that the receiver has a built-in preamplifier). A degree of redundancy may therefore be necessary if equipment faults are not to put the groundstation-out of action for some time when a fault occurs. If equipment is donated there may be the problem of incorrect voltages to consider. h) Interference from other local sources of radio emission may have negative effects on reception, particularly in industrialised or urban areas. While PTTs should pro- tect against this the regulations may be lax or difficult to enforce. i) It will be realised that any individual with the relevant equipment can pick up the data being broadcast from the satellite (HealthSat). To date no privacy has been offered on communications via satellite. This has been a cause of concern and some individuals or organisations are naturally unwilling to use the system until confidentiality between the originator and receiver can be guaranteed. Adding a password to standard compression software (eg. PKZIP) is clearly one possible method (although the password must then be transmitted by some other means) but something more definitive may need to be devised for long-term use.

J) The link between the messages to and from the satellite and to and from the terrestrial locations is still in its infancy. Communication with the satellite itself is based on a modified X.25 protocol with each message having a large header record containing transmitting and receiving addresses, keyword and compression information, as well as the message number. This header needs to be added to the message before transmission and stripped off after reception.

Since FidoNet is used for moving the messages about within Zambia, and that has its own message protocol, each message from the satellite currently has to be loaded in to the FrontDoor fido software 'by hand' and have an appropriate fido.address added to it. The satellite software will by then have been unloaded and the fido software loaded so that the message can be sent off to the fido email system via a standard modem.

The software for control of the groundstation has improved to the extent that it can basically run unmanned (sending any waiting messages and picking up any locally addressed messages from the satellite automatically). It is clearly now desirable that the software is produced (and it is being commissioned) which will

33 automate the procedure such that once a satellite pass has taken place newly arrived messages will automatically be put in fido format, readdressed and sent off to the terrestrial system. Likewise before each pass incoming messages should be picked up, reformatted and made ready for transmission. k) Calculation of satellite positions and timings is done using a piece of software which currently has to be run manually since it is interactive (and requires a maths co-processor due to its complexity). Coordinates are then loaded into memory and a TSR looks after the interface with the transceiver and causes automatic tuning every few seconds while the satellite is overhead. Again this software will ultimately need automat- ing if regular manual intervention is to be cut out.

1) Any fido network based on the local telephone system is likely to be subject to a number of problems, however cooperative the local PTT (with whom good relations are maintained in Zambia). These will include difficulty in obtaining an additional dedicated line where considerable patience may be required. (Additional difficulties were experienced at the University Teaching Hospital when a wire needed to pass through the mortuary!). Lines may also be susceptible to disturbance during the rainy season or when interrupted by normal PTT problems such as cable disturbance or theft. Non-payment, international or otherwise may result in unnotified cutting- off of the service. m) Obtaining the backing of the PTT is clearly essential to the whole HealthNet operation. From initial acquisition of the licence and,ensuring freedom from interference, up to installation of new lines; checking of line availability to each hospital or rural site, and maintaining of links once in place; discussion of best and most reliable lines of communication within the country, as well as discussion of future PTT plans, the PTT have to be involved. It will be noted that there is currently no 70.25 system in place for data communication in place in Zambia although the PTT have planned for its introduction in the future. Links within the country also vary greatly, from good digital exchanges to old electro-mechanical ones; from modern line links to radio-links with outlying areas. These have not yet all been tested for fidonet reliability. In addition, difficulty has been experienced with a particular batch of 25 slightly defective modems purchased to create the fido network. Although a fault was quickly diagnosed, getting a manufacturer in Canada to accept responsibility and to take action is very difficult with no leverage over them and at a distance of 10,000 km.

34 4.2 Human Problems a) Zambia has been fortunate to have a set of skilled and dedicated staff at the Computer Centre who were able to make a new initiative work. Such skills are not always available and once staff are trained up in such new areas they become even more marketable in an economy where they may be justifiably tempted away out of the public sector or indeed out of the country. Staff training is a further problem and staff retention must be an integral part of any project drawn up within the African context. b) The international radio 'ham' community can clearly be useful in assisting with the setting up of a packet radio scheme as they will have the training and authority to use systems on at least an amateur basis. Some African countries do not allow amateur radio. Others do, but hams may be rare. c) Budgeting for ongoing equipment needs or resources to run a communications network can be hard in economies where exchange rates and general income can be highly unpredictable. This is likely to be particularly so in an institution like a University which relies on government grants for its running. Some equipment will require hard currency and that again may be difficult to pbtain or require the use of a complex system of application for its acquisition. d) In a situation of relative isolation it may be hard for staff operating emerging systems to know where to turn to for help or what facilities may be available through other systems elsewhere. They may not know what they are missing or what they should be aiming at. This may make motivation hard and finding the required skills level even harder. This is the more so because much of what is taking place in the communication field is new. and many people are working on many new projects without coordination. It is thus quite likely that someone may have software or technology to suit an application without other potential beneficiaries being aware of it. e) One other effect of the current isolation experienced by some African (and other technically developing country) sites is that the potential user community may not be aware of the benefits that can accrue from the use of modern communications systems or of the levels of technology required to implement them.

This leads to a number of immediate side-effects. The first is that the take-up on a system when it is first put in may be slow. People have become so used to not communicating, or to doing so by slow conventional means, that they will not easily begin to use anything

35 They may also not be frequent computer users at cne present time so the leap may be a double one. There is also a belief to overcome that if it is difficult (as it often is) to phone across town, then international emai-1 communication must be impossible. They must quite reasonably see proof, and experimentation must be an integral part of any proposals being put forward.

There is also skepticism that any system could provide free (or at least very cheap) communication: such ideas are normally seen to be short lived. Additionally it is common to perceive an attitude that such technologies as packet-radio or fidonet are inappropriate for the current status of development within a country. Some aid donors in particular are yet to be convinced that communication or information provision should take a high priority in developmental terms, preferring instead to put money into more traditional areas such as agricultural training or direct medical treatment. The argument goes that if a hospital cannot afford Aspirin should it afford a computer? The argument may be spe- cious but must be won.

The concept of sensitising the targeted user community to the benefits of new communications systems is thus an important one which must not be overlooked at the planning stage. The University of Zambia has been through an experience of seeing many people calling in each day to collect messages on the email system but noting that noone was actually initiating mail. There were thus no replies to collect and the system began operating in something of a vacuum! Once users began to send messages and make the world aware of their presence then messages begin to flow quickly. f Some problems are related to those normally experienced within regular situations and require only persever- ance. A system may fail because there is a paper short- age and no messages are getting printed. A plug may be required and not available. Wires may have become detached and no one is there to notice. A telephone bill may have gone unpaid and the system stops working without attention being drawn to it. A modem may have become inoperable at one hospital and there is no transport available to get there within the next week.

The role of the human factor is thus high, and if a system is to be sustainable the human element must be properly considered. The right (highly motivated) people must be in place with the right backup and support.

36 5. The Future

Zambia's experience with HealthNet is still in its infancy. A lot has been learned but there is still a long way to go before a wide base of medical practitioners is benefiting from the attempt to reduce their current isolation.

The experience to date though has been a positive one. Although there have been many technical obstacles to overcome there has been generosity on behalf of the service providers and patience on behalf of the user community who from the most senior to the most junior await the arrival of a more fully operational system with great interest and enthusiasm. It is to be hoped that as the technical problems are overcome a very real human benefit can be obtained and health become a more readily available product where it is so badly needed.

July 1992

37 HealthNet Satellite Communications System

'Stare-and-forward' Satellite

W 00

Worldwide Locations HealthNet Satellite Groundstations in Africa Proposed HealthNet in Zambia

0 Soiwezi T9 Kasama Ndola TDRC

Ch ipata UTH IlUnza

of HF

ivingstone - District 0 Hospital EASTERN AND SOUTHERN AFRICA NETWORK (ESANET)

COMMUNICATIONS RESEARCH FOR DEVELOPMENT

THE EXPERIENCE OF THE NAIROBI NODE

Prof. Anthony J. Rodrigues Dr. Geoffrey M. Macharia Institute of Computer Science University of Nairobi P.O. Box 30197 Nairobi, Kenya.

1. PROJECT GOALS AND OBJECTIVES

The East and Southern African NETwork (ESANET) is a research project, funded by the International Development Research Centre (IDRC) of Canada, aimed at investigating various microcomputer based methodologies for communications. The countries of the five participating institutions are all members of the Preferential Trade Agreement (PTA) region. These institutions are the Institutes of Computer Science at the University of Nairobi and Makerere University, and the Computing Centres at the University of Dar es Salaam, the University of Zambia and the University of Zimbabwe. The general goals of the project are:- a. to experiment with microcomputer based communication networking in order to acquire the necessary technological capacity, and, b. to promote more effective and efficient communication within the research community in the region.

Within the scope of these general goals, the specific objectives of the project are :- i) to experiment with alternative modalities and techniques for data communications among and within the five nodes (Nairobi, Dar es Salaam, Lusaka, Makerere and Harare). ii) to evaluate the technical, economic, sociological and management aspects of the communication network experiments; iii) to disseminate information to the research community within the region about the development and the results of the project with a view to increasing the awareness of possibilities, stimulating new and wider applications, and inviting feed-back on related topics; and iv) to make recommendations to the research community (users and institutions) and telecommunications authorities in the region on cost effective data communication modalities, and appropriate

41 network models and policies for specific environments applications. and

The emphasis was on microcomputer based communications systems. Thus the project had no intention of offering private data communication network services in competition to any commercial services already offered by the respective Posts, Telegraph and Telecommunication authorities (PTT) in the participating countries. In fact one entire experimentation strategy uses dial-up modems and while the public network, only the Satellite/ground-station component of experimentation the other strategy is private, other linkages being public. The satellite based strategy was primarily intended to facilitate Medical Research. Its usage, the users and the data volumes available, restricted, are are not-for-profit and are non-commercial. It was important that the developmental impact in the field of Health Care and Communications be weighed positively against the perceived revenue loss of for the PTT authorities. A convenient and effective means to monitor the usage of the network for security was identified. purposes by the KP&TC

2. COMMUNICATION METHODOLOGIES

Two methodologies were proposed based on different communications media to facilitate research and experimentation:-

i) Public voice telephone lines, and ii) Packet Radio Satellite.

The only private component of the second strategy was the transceiver- satellite linkage. As such, and given the strict restrictions on the usage, the users and the data volumes as outlined above, the issue, raised by the regulating section of the KP&TC, of ESANET providing a private data network, was placed 'in its proper perspective. 2.1 CONVENTIONAL PUBLIC TELEPHONE NETWORKS

This strategy uses the Public voice telephone lines in conjunction with type-approved data modems and a microcomputer-based communications software package called FrontDoor developed by the Association for Progressive Communications (APC).'The APC network comprises some seven major nodes around the world. Kenya is set to become a node in this international network.

The ESANET network is in the public domain and in addition to the five Institutes/Centres of Computer Science in the region, the users are connected to the network with the appropriate public domain technology viz. type-approved modems, etc. The information communicated is expected to be similar to that in the public network. 2.2 PACKET RADIO SATELLITE SYSTEM Use of the Packet Radio Satellite methodology, though private, involves microcomputer based communications software packages, a terminal node controller (TNC), and a packet radio UHF receiver and VHF transmitter set at each 'ground' station for communication with a low-earth orbit data communication satellite using either an omni-

42 directional whip antenna or a uni-directional Yagi-type antenna. Thi; facility does not offer on-line communications capability; it is base on the store and forward principle (following the bulletin-board modu, operandi).

2.2.1 Communication Packages

The communication packages in the ground station for spacecraft tracking, station control, and satellite monitoring are:-

GrafTrak II for real-time display of the simulated path and the coverage area of a selected satellite relative to a fixec observer,

Silicon Ephemeris (Seph), a program that calculates the positior of the satellite by knowing its ephemeris, the time of the day and the geographical location of the ground station. This information is then used both to calculate the satellite viewing angle, and to provide input to the Kansas City Tracker.

Kansas City Tracker which, using information generated by Silicon Ephemeris, may perform any of two functions:

a) feed frequency correction data to the transceiver. b) where applicable, provide steering control for the Yagi-type directional antenna.

PACSAT ground station software suite which includes a module for capturing broadcast messages from the satellite and a module for ground station functions which include compiling the up-link messages and un-packing the down-load messages. 2.2.2 Ownershic and Manaaement This technology has been made available by SatelLife which is a non- profit, non-commercial organisation whose goals are to facilitate Medical research in developing countries. This organisation buys time on satellites, the current one being UOSAT-3 developed by the University of Surrey. Owing to payload restrictions, the total volume of up-linked and down-linked data traffic available for each ground station is limited to 500 pages per month per station for not only the Medical Faculties, but also the participating Institutes and Centres of Computer Science from Universities in the region. 2.2.3 Tyre of Information In addition to the medical information exchanged by the Medical Faculties, the information exchanged by the Computer Science Institutes and Centres will be non-commercial, not-for-profit and non- classified, and will be strictly limited to experimental data and project documentation.

2.2.4 Security

Information for up-loading to the satellite is received at the `ground' station either by hand-delivery, or via the public network strategy outlined above. Similarly, down-linked information, which is received in broadcast mode, will be passed on to the destination by 43 use of the same public network modality or by hand-delivery. therefore This affords a convenient method for monitoring of the ground station traffic for security purposes.

3. CONNECTION TO THE PUBLIC NETWORK

In Nairobi, two telephone lines have been installed. They facilitate, on one hand, greater telephone line reliability and availability to ensure effective utilisation of the public domain communication strategy, and on the other hand, efficient communication information of between the ground station, at the Institute of Computer Science, and the Medical Faculties. The earth connected station is not yet to the public network: the file transfer is manual at present. Technical and regulatory issues are yet of automatic file transfer. to resolve the issue The data modems are used exclusively, after type approval, public with the domain network above for the purposes of approval experimentation. Type for the modems has been sought by the four other institutions from their respective posts and telecommunications authorities PTA region. in the

4. SECURITY: INFORMATION FLOW MONITORING ON ESANET BY KP&TC

(a) The first methodology involves the existing telecommunication infrastructure using dial-up modems, microcomputers and a software package called Frontdoor. Monitoring of information flow in such a network very easily can be achieved by making the KP&TC a network 'point'. This requires the KP&TC to provide a AT- Compatible microcomputer on which the software would be installed and one person trained in its use.

(b) The second and complementary strategy is to integrate data transmitted through low orbit satellite communication technology into the network at Nodes, involving VHF/UHF transceivers and associated accessories.

The data flow can be accessed in the following two ways:

M Radio-link at the KP&TC monitoring all up-loaded and downloaded data into the Node at the Institute of Computer Science.

(ii) A special dialling port into the Node to monitor the data using Frontdoor with KP&TC as a privileged user. This option which provides effective vetting is technically feasible immediately. i. NON-GOVERNMENTAL ORGANISATION (NGO) INVOLVEMENT 'wo of the NGO' s involved in ESANET since its inception are Sate 1 L i f e ind VITA, with SatelLife playing a major role. SatelLife is an rganization set up by the International Physicians for the Prevention >f Nuclear War, and Volunteers in Technical Assistance (VITA) is a rivate, non profit voluntary organization established by scientists ind engineers for the purpose of sharing their knowledge skills and

44 experience with people in developing countries. FrontDoor, a software package which features networking a comprehensive interface was developed by an non-commercial NGO the Association network, of Progressive Communications (APC) which has seven countries; nodes in Canada, USA, UK, Sweden, Australia, Brazil and Nicaragua. These nodes exchange electronic material mail and conference and are in turn connected to JANET and BITNET virtually all to which Universities in the UK and the USA are linked, respectively. Thus the APC provides an opportunity African Universities for:-users in in ESANET to establish regional and international contacts. ELCI, the Environment Liasion Centre in training and Nairobi'- provides expertise in the use of FrontDoor to ESANET, During the initial users. phase this was provided on a goodwill basis, however, this is now provided on a varying commercial basis with the rates stipulated in US dollars.

6. INTERNATIONAL DATABASE ACCESS Through the use of Frontdoor which accesses the large variety various APC nodes, a of international databases for academic usage available. Also become SatelLife and VITA enable access to several and European databases American to carry out medical literature search services such as Mediine, abstract and bibliographic services references and document for selected full text medical articles. will connect In due course this to networks of academic institutions such BITNET (USA) as JANET (UK), and EARN (Europe). However such database to be achieved. accesses are yet

7. PROJECT ACTIVITIES 7.1 START-UP-MEETING

The project start-up meeting was held at the Institute Science, Nairobi, of Computer on 21-22nd November 1990, to map out direction, establish the strategic and prioritise activities, and to plan and schedule the first phase of the project. After reviewing the communications region, and networking developments in the as received by verbal reports from the five participants, and principal the presentations on FIDONET and FrontDoor which use the existing telecommunications infrastructure, satellite and on packet-radio based communication system, it was communications agreed to adopt these two systems. The aim was to use these and any other appropriate communication methodologies to fulfil specific the general and project objectives of ESANET as listed above. The results of this initial meeting were:- i. tentative network topologies for both FrontDoor and packet-radio based communication methodologies, ii, hardware, software and staff requirements for implementation, project iii. project implementation schedule, iv, provisional agenda for the design workshop in Harare, v. list of prospective users, and

45 vi, list of the initial activities.

7.2 DESIGN WORKSHOP

A design workshop was held in the Computer Centre of the University Zimbabwe from of 26th to 30th May 1991. The objectives were: the project's progress; to review to receive country reports; to hands-on training carry out on both the FrontDoor electronic mail management system and the packet-radio satellite develop based communication system; to a strategy for incorporation of develop further user groups; to project evaluation plans; to design and co-ordinate experimentation phase; the and, finally to review the project schedule. Each of the country reports outlined each ESANET node's.progress in: applying for telephone lines; obtaining obtaining type approval of modems; licenses for for use of packet radio systems; installation of hardware and software; and the response from groups. prospective user

The principal the design workshop were:- i. hands-on training for system operators on an updated mail electronic management and conferencing software suite. This included the GoldED message editor, the IMMAIL mail and conference suite and a new version of FrontDoor, ii.. a review of ESANET topology, iii. globally unique addresses and addressing modes for node each ESANET and prospective users in readiness for attaining full FIDONET node status, iv. deqision to establish electronic mail-boxes in GREENNET, London, for communicating between the ESANET nodes, v. experimental format and details, vi. proposal to commission accounting software for incorporation into each node, and vii. list of activities for the experimentation phase. The comprehensive report of the design workshop country incorporated: the reports; the network topology; the FIDONET addresses; the names of the GREENNET mailboxes for each node; outgoing the addressing mode for and incoming mail; and the templates for structuring and standardising the experimentation.

7.3 PROCUREMENT OF HARDWARE AND SOFTWARE The Packet Radio Satellite technology as outlined in section 2.2 has been made available by SatelLife whose Medical goals are to facilitate research in developing countries. This organisation owns and manages time on communications 'micro'-satellites, the current being one UOSAT-3. Owing to payload restrictions, the total volume of up- linked and down-linked data traffic available for each ground station is limited to 500 pages per month per station for not only the Medical Faculties, but also the participating Institutes and Centres of Computer Science from Universities in the region. The hardware and software equipment specifications were defined at the start-up meeting held in Nairobi on 21-22nd November 1990. On advice

46 from an expert based in ELCI, Nairobi, it was decided to procure the equipment from NIRVCENTRE, an NGO based in Toronto, Canada, due to the cheap prices offered. The order was duly placed with them on 5th April 1991. However, the project equipment was delayed for a number of months due to various reasons, the principal one being NIRVCENTRE's inability to provide the equipment exactly as specified; particularly with regard to input voltage ratings. To avert further delay, they were authorised by the Project Manager to ship the next best available alternatives. Thus Nairobi received their consignment in September 1991, Lusaka in October 1991, Kampala in February 1992, Dar es Salaam and Harare in April 1992! This delay in delivery of project equipment, ranging from five months to twelve months, has correspondingly delayed the experimentation phase of the ESANET project. Furthermore, the non- compliance with input voltage ratings, especially in the case of the external modems and printers, has caused additional delays and unbudgeted expenses in commissioning the equipment. The reliability of the equipment for end-users borders on the unacceptable. The cheap equipment prices obtained thus do not appear to compensate adequately for these operational delays.

8. CURRENT PROJECT STATUS

8.1 THE ELECTRONIC MAIL SYSTEM

It has been found that making inter-node calls is difficult, primarily due to poor telecommunication infrastructure for calling between countries within the ESANET region. Making calls outside Africa however, is relatively easy due to configurations of international communication satellites. Therefore it has proved expedient to set up electronic mail-boxes for each ESANET node on GREENNET in London, U.K. which is a node of the APC global network. These mail-boxes have facilitated communication between ESANET nodes while awaiting project equipment to be shipped. The original plan, yet to be implemented, called for upgrading of one of ESANET's nodes into a full APC node.

In Nairobi, the Institute of Computer Science was, until March 1992, connected to an old relay-type telephone exchange in Kileleshwa which led to severe line reliability and availability problems. However, the Institute is now connected to a modern digital exchange in Westlands which provides more efficient services. These services are expected to improve further, hopefully in the near future, once the few hundred trunk lines between this new exchange and the central switching office are increased to a more reasonable number. The Nairobi node for conventional modem communication has been experiencing some installation glitches. The hardware supplied has broken down time and again. Being unable to guarantee that the system is on 24 hours a day, f u l l node f a c i l i t i e s have not been offered to the user community. Furthermore, the external modems are not usable as they are since they need a 240/110V transformer which was not provided by NIRVCENTRE. Thus providing these modems to the prospective user base will entail additional unbudgeted outlays for stand-alone 220V/110V ac transformers. This illustrates once again the trade-off between equipment cost and appropriateness.

47 8.2 THE PACKET RADIO SYSTEM

In January 1991, a complete ground station was successfully installed in the ICS, Nairobi, for testing and demonstration purposes. This exercise not only enabled personnel to be trained in the installation and operation of a packet satellite ground station, but also facilitated in the application process with the KP&TC. Further to this a request was made, in August 1991, to the Government of Kenya by the University of Nairobi for authorisation for the use of the satellite communications system. This approval was granted in November 1991. A further application for approval and allocation of frequencies was made by the University to the KP&TC. The approvals were granted in June 1992 subject to a type approval fee of Kshs. 2970 (US$ 90) and an annual licence fee of Kshs. 6364 (US$ 193).

The approval is conditional on:

i) the equipment being operated on the frequencies approved, ii) the system being operated on a "Secondary" basis, i.e. it may suffer interference from existing services and shall cease operation should it cause interference to other existing service, iii) the system being used for research purposes only and on a non- commercial basis.

Meanwhile, the University of Zambia and Makerere University, Uganda, have received licenses to install and operate ground stations for the packet radio system. After receiving radio equipment from SatelLife in June 1991, the Zambia ground station is now fully operational. . 9. FUTURE PLANS

The attached Gantt chart in the appendix gives the activity and timing schedule as planned. However, delays in procurement and full installation of the electronic mail facilities as noted earlier have upset this schedule. An attempt to get the project back on track was to have an interim workshop in Zambia in April 1992. This was agreed on at the design workshop in Harare in May 1991. After serious consultations between the node co-ordinators and the donor, it was agreed to revise the project completion date from 31st October 1992 to 31st October 1993. This will enable the interim workshop to be held in Lusaka in early November 1992 with the evaluation workshop at Arusha re-scheduled for mid-August 1993.

It is hoped that all the nodes would have; obtained licenses for installation and operation of packet satellite ground stations; received, installed and tested all the project equipment, both for the telecommunication'based methodology and for the ground station; obtained type approval for the modems supplied; and, have had some experience in using the electronic mail management system. This meeting will facilitate the consolidation of the experimentation phase and thereby assist in the evaluation phase.

Once the respective PTT licensing formalities are concluded and the project equipment is fully installed, it is expected that ESANET will attain full operational status very quickly thus enabling the experimentation and testing phase.

48 10. CONCLUSION

Although this is a research project with obvious technical and economic components the more challenging aspects that appear to emerge during the course of the project are sociological and management issues. Thus the factors that seem critical to the project's success are, on one hand, the key actors involved, and on the other hand, their perceptions of both the project and their respective roles in the project vis-a-vis its objectives. Thus the realisation of the project goals and objectives entails effective management of the conflicting and complementary interests arising from these various perspectives. A detailed case study with these issues in mind may shed light on the delicate and complex interrelationships at play between the key actors.

49 Digital Radio Technology and it's Application in Tanzania. International Workshop on Digital Radio Technology and Application in Nairobi Kenya =August 24-26,1992 Presented by: William Sangiwa. Overview.

Due to the poor existing communication infrastructure, that is telephone/fax, telex and postal services. The use of Radio communications in Tanzania has been going on for sometime now as a means of reliable communications and this has been only in voice communication and not in data communications. The use of digital radio technology in data communication has been a new experience altogether, although there has been some researches done on the issue, none of them have came out of the research phrase. The introduction of digital radio communication on data communication by the NealthNet has been one of it's kind to go beyond the research phrase into public use. The use of this technology will grza y help in receiving and dissemination of information to the places which usually receives or sends very little, most of the time it is outdated or does not send or receive at all.

Although the flow of information is mainly from developed countries to developing countries with little input from developing counvies, it is prodictcd that, with the communication facilities that would enable to have an affordable, reliable and considerably fast way of moving the information, the flow of information will be bidirectional from north (developed countries). to south

(developing countries) and vice versa

In Tanzania, the current flow of information is greatly influenced by political, economical and existing telecommunications infrastructure Although since it's installation there is no change so far (as we don't expect rapid changes) on the flow of information, many of the users or rather would-be users arc optimistic if it will be sustainable means of communications.

50 In the developing countries, Tanzania in particular, politics have lead to strong censorship to the means of communications or rather the information which goes either in or out of the country.

Most of the time the means of communications (information) are controlled by government owned agencies/companies which have the monopoly and this has played a big part in slowing down the developing of communications infrastructure as they had no competition.

Long delays in waiting for the response from the operator to connect you for an international call most of time in the range of 30 minutes or more as most of the lines needs operators assistance for an International call and the cost of an international call puts the whole idea of information exchange at bay. The cost of making an international voice call in Tanzania is very high compared to the income of the most of Tanzanian's ( USS 7.00 per minute while the highest paid civil servant is paid approx. USS 100 ) and this was aimed at reducing the number of outgoing calls and thus making the Tanzania PTT paying little for the International calls to other PTT and thus saving the foreign exchange.

There is also a feeling of ownership of the facility, although it has been emphasised on several occasions that this facility is free for use to all user's. Many still strongly believe that one day they will be axed off the system, and all their investments either in terms of time or any other-way will be completely gone. This tendency is attributed by earlier experiences they had. The issue of confidentiality has also been raised at some point, although they had been some assurance that the chance that the message will fall into wrong hands are minimal still the users/wouki-be users

arc sceptical on the issue. Despite that, there is general lack of information, the introduction of

this technology has received very little attention to the research and teaching communities. For

example most of the researchers or lecturer or students want to see something like online

database terminal for medical searches and literature review, they want to see instant replies with

turn-around time of few minutes, this is because some of them have studied abroad and have

come across such facilities, even though for the time being it takes months to get the same results,

if your lucky to have a friend on the other side of the world [north], most of the proposed

researches and papers arc country based, that means there is little or no sharing of ideas before

51 or during the preparations and executions of any project with people outside the country who had or are experiencing the same problems besides the donors (if they happen to be foreign donors) until it's publication.

Organization Involvement

As part of a project funded by IDRC-Canada to link all universities of East and Southern Africa

The idea of using Satellite to link the universities was used as one of the options. My organization, Muhimbili Medical Centre, incorporating the Muhimbili University College of

Health Sciences of the University of Dar Es Salaam together with SatelLife an organization based in Boston, USA , are trying to fink all health personnel of Tanzania with others throughout the world through the Radio-Satellite link and make the access of health information more easier,

faster and reliable. At the medical library, the Satellite ground station is housed and this will be

the point of sending and receiving the informatk-a to and from Tanzania It has been chosen to

stay in the Library as this will allow the station to be accessible to everyone. The accessibility of

the system referred here is the services that are offered by the station not physical operation of

the station. So far there is very little information that has been offered to community on the

existence of such a facility. Besides allowing the it's staff to be on the project (system operator's). the There is very little assistatxr offered to the project From the experience gained during

installation of the system it has beat clearly observed that, the involvement of the

administrators/managers in planning and execution will help to have ready made influence on

them and thus gaining organizational support

Future Application of the Technology

burden With more stations joining the HcalthNct The existence of the HealthNet will easy the at the right that is currently carried by the health professionals in acquiring the right information

fax, normal time. For example most of health professionals arc out of touch with telephone or time the cost are post mail takes on ages, sometimes it will never reach there at all. Most of the 52 met by the callers or senders, taking into consideration the low income they earn and the high cost of making a call, it is highly unluckily to have constant communications. As the HealthNet has proved, with affordable installations and easy to use systems that will provide reliable and sustainable way of communications, the future of this digital radio technology especially in developing countries is very bright. The stations have proved to be easy to mount and operate, maintenance free and incase of breakdown they can be serviceable in developing countries. Views.

As it has been the case with the HealthNet in Tanzania, it is clearly that most of the success that are enjoyed with the HealthNet arc purely by personal initiatives or convictions of few persons.

This has showed that before attempting any further expansion it will be necessary to establish the personnel that will be involved with commitment and willingness to work. As the cost of living are sky rocketing it is highly unluckily to get a free hand to work with the projects. Generally, speaking as this technology is very new to third w irld, there is also a need for training operators

how to use the system otherwise most of the installation will run aground, as there will be no one to operate them Here I'm suggesting something like training of a trainer who will be incharge of

local training. Involvement of administrators from the beginning and not only signing of the contracts and letters will help in making them feel as part of the project rather than just invited

guests. This will have an impact on the overall organization involvement as they are the one on

the decision making levels who usually consider projects as personal achievements rather than

dcYclopmcntal achievement to their organizations

53 Experiences in Setting Up a PACSAT Station in Tanzania

H.R. Mgombelo and S.J. Braithwaite

Abstract

Experience and details of setting up a PACSAT message store-and-forward ground station in Dar es Salaam. Tanzania, are described. A notable feature of the station is that it has a low operator maintenance requirement since accurate knowledge of time of day and satellite orbital elements are not required.

Introduction

In the summer of 1991 a British Council academic link was formed between the Universities of Dar es Salaam and Southampton in the field of Electronic Engineering, specifically in the field of communications (1]. The link fosters collaborative research, in this case mainly in Rural Telecommunications.

The authors particularly wanted to maximise the integration of the departments in this research and it was apparent that the grade of service and expense of conventional international telephone connections was inadequate. There were requirements for batch access to computers in Southampton for simulations using parallel compubers a need for research reports and discussion documents to be passed between the institutions. .

54 The VITA organistion at the same time advertised for project proposals under their VITASAT initiative using the Uosat 3 (also known as U014) spacecraft. The project proposal outlined above was accepted and we set about planning for the installation of a ground station in Dar es Salaam [2]. A satellite station already existed in Southampton that we could use.

This paper address the experiences we have had in setting up the satellite link and we describe the ground station equipment in Dar.

Station Mission

There were three main purposes of installing a PACSAT satellite station at the University in Dar es Salaam. Firstly, the collaborative research between the two Departments involved indicated a need for researchers in Dar to run programs on large Transputer arrays in Southampton as part of a research programme on rural telecomms networks. Such facilities are not available in Dar. We realised that this would make an attractive demonstration of a PACSAT link through the VITA demonstration programme as well as meeting a perceived need. VITA adopted the project on the strength of this proposal.

Secondly, this concept of remote batch processing will be extended to another need we foresee next year (1993) for relaying radio propagation measurement data from field measurements in Tanzania to Southampton for processing. This will involve a transportable PACSAT ground station.

Thirdly, the easy data communications that PACSAT provides can meet the need we have of efficiently transferring large research reports, specifications, plans and inter-group communications.

As a spin-off for the students in Dar, the station provides a valuable piece of working data communications equipment that can be used as a basis for undergraduate project work and study. This is indeed already the case with the satellite station in Southampton.

55 Equipment

The funds available for buying equipment for the station were fairly limited. We made a decision to build as much of the equipment as possible and opted for a low-complexity solution, trading off grade of service for cost [3].

An IBM PS/2 /30 was available for pressing into service as a part-time station controller and we bought Pac-Com Tiny-2 terminal node controllers and modems.

Rather than go the route of a fairly expensive commercial transceiver ,

costing some £1300 , we opted to build transmitter and receiver units from Wood and Douglas (W&D) amateur radio kits and modify them for use on the 9600b/s narrow-band FM transmissions used on the U014/UO22 PACSATS. The transmitter is a single-frequency unit generating about 25W output, running from a 12V supply. The receiver is similarly a single-frequency unit powered from 12V and uses an IF processor add-on board designed in the satellite labs at the University of Surrey. The IF processor provides automatic frequency search-and-track over about 17 kHz and by-passes the W&D audio output stages to provide a suitable connection to the TNC. 12V Power supplies are provided by laboratory bench supplies.

We opted for a low-gain, non-steerable antenna approach for the following reasons: Firstly the cost, maintenance and operation of a non- steering antenna is significantly less than that for a steerable arrangement

(materials costs were some 1 /10 of a rotator-mounted antenna set if the associated PC hardware and software are included). With automatic frequency tracking and non-steerable antennas, no maintenance of Keplerian elements or time-of-day is necessary. This is not currently the case with stations that use steerable antennas or a commercial amatuer radio transceiver. We built the 3- turn helical antenna chosen for the down-link, and used a standard 5-element crossed Yagi (with two sets of elements removed to lower the gain) for the 140 MHz transmit antenna. Secondly, the cost to grade of service of using a low- gain antenna (roughly half of the available passes are usable) seemed reasonable for accessing the satellite. (There has been no noticeable congestion of the satellite when it is in view of Dar es Salaam). A-0.5 dB no 56 figure low-noise amplifier is mounted on the ground plane of the helical antenna.

The antennas are pointed toward the south (there is a rise in the ground just to the North). The link is workable over a little more than half the area of the sky, giving some 10 minutes per pass maximum usage (typically seven minutes).

We have recently prepared for trial a pair of 'Eggbeater' antennas for Landrover-based transportable operation and will report on them in the conference session.

Software

All PACSAT operations up until July 1992 were on the amateur band frequencies and were consequenby limited to station work-up, faultfinding and trial communications. We were therefore in a position to use the Surrey University software suite PBPG. Recently we have worked with the VITA software suite which provides more advanced call logging, though in its current version operates in connected mode only, giving a reduced throughput compared to Broadcast mode operation.

Licensing

Perhaps the most frustrating element of the last year of setting up the PACSAT link has been that of licensing. We have encountered difficulties in both the UK and in Tanzania. The UK situation is that there is no recognised satellite allocation in the frequencies used by the VITA payload on UOSAT 14. Consequently, there is reluctance to move toward licensing any kind of use on these frequencies.

As a backup solution we intend to use the US VITA station as a gateway to the Southampton facility, with messages being transferred onto FIDONET

57 and sent over the Atlantic. This, of course, is somewhat short of our original mission. We are continuing to investigate the possibility of UK licensing.

As far as the Tanzanian side is concerned, the problem has been the long process required for a licence to be granted. Application forms have first to be processed by the Tanzanian Post and Telecommunications Corporation (TPTC). They check for spectrum allocation difficulties, among other, unknown tasks. The forms are then passed on to the Ministry of Communications after whose approval are then passed back to the applicant through TPTC. Recent changes in the structure of TPTC may have contributed to the long delay of 10 months (so far) of obtaining a licence to operate on the VITA frequencies. It has been noted that our experience here is not uncommon in other developing countries.

Training

The station in Dar is operated by staff of the Electrical Engineering Department. There is an obvious need for any operator to be computer-literate, which is widespread amongst the staff. A half-day training session (due to limits of time after a lengthy delay with. the in customs) was held for half a dozen of the staff. Some six months of good operation of the equipment followed. The Southampton amateur station was operated by one of the co- authors and amateur-qualified students.

Operational Experience

We have mentioned above our expected communication time per pass of some six minutes with non-steerable antennas. This is sufficient to receive directory information and download mail destined for the station in Dar. In the first year of operation mainly test messages were sent and received using the amateur payload on U014.

58 In early 1992 there was a switch to U022 for amateur use, with UOSAT-14 being used entirely for non-amateur communications. This somewhat left-footed us as it required the transmitter and receivers to be re- crystalled for the new satellite (showing an obvious advantage for the commercial transceivers). Rather than perform the change in Dar, we waited until the sets could be returned to Southampton for re-alignment, which was duly done in March 1992.

Following our opening ceremony for the station in July 1991 and its subsequent coverage in the national English language newspaper there has been a widespread interest in the station and the potential of messaging via PACSAT. The press reports were subsequently carried by Africa-oriented communications trade magazines. Several requests have been made for occasional use of the station to convey data to Europe and the USA. This interest has extended to the investigation of new VITA projects elsewhere in the country.

Healthnet

There is another PACSAT station in the vicinity of the University station in Dar. The Satellife Healthnet project have a steerable antenna ground station about 15 km away from the University at the Muhimbili University College of Health Sciences. The two stations will be using the same frequencies on the satellite. With the packet multiple-access nature of the link to the satellite, this is not expected to be a problem, especially when broadcast-mode protocols are used. There is free exchange of information and experience between the two stations which provides much appreciated contact.

Future plans

In many ways our use of PACSAT for academic collaboration is still in its infancy. There are several trials and improvements planned:

59 From March next year we expect to use the station as a transportable one mounted in a 4-wheel drive vehicle to help us in radio propagation measurements in remote areas. We plan to use the mz tStarater antennas for their compactness and transportability.

Work has been progressing at Southampton on a UNIX-based server for PACSAT operations, both as an archiver for Amateur use and as a server for remote batch processing. We intend to complete this part of the project as a demonstrator for wider applications such as mapping, prospecting, weather prediction and other applications requiring the type of computing power not available (or desirable) in a remote situation.

The question of whether to install steerable antennas is under discussion. If the accessibility of the satellite becomes a difficulty because of increased loading from other stations within its footprint then the step to parity of EIRP and receive gain with other stations may need to be taken. In the meantime, we enjoy the freedom from needing accurate time-of-day and up-to- date orbital elements.

We mentioned above the interest in relaying information for other interested parties. We expect this to become an increasingly common request. We have yet to establish the grouted rules with VITA for agreeing to act as an information gateway in this way.

Conclusions

In summary, the experience of setting up the station has been an enjoyable and interesting experience. Although there have been several hurdles to jump the fruitful results and the feeling of being in contcat with the rest of the world and with collaboratiog institutions has been rewarding.

We would like to extend our appreciation to all who have assisted us in this venture. Special mention goes to the British Council, VITA, and the University of Surrey Centre for Satellite Engineering Research.

60 References

[1] "Link memorandum for collaboration between the University of Dar es Salaam and the University of Southampton " July 1991. British Council, Manchester.

[2) "Pacsat communications experiment (PCE) demonstration memorandum of understanding between VITA and the University of Southampton." June 1990. VITA. Arlington, Virginia.

(3] "PACSAT Station local documentation" July 1990. Dept of Electrical Engineering, University of Dar es Salaam, Tanzania.

Contact Addresses: (S.J. Bradwaite) Univoers4 d Southampton. Department of Electronics and Computer Science, Southampton, SOO SNH. Tel +44 (0)703 5133825. Fax (0)703 592885. E-mail sjb®ukac.soton.ecs. G7JMK

(H.R. Mgombelo) Unwsrty of Dar es Salaam. Dept of Electrical Engineering, PO Box 35131, DSM, Tanzania. Tel +255 51 41968 Fax +255 51 48802 5H3FE

61 SEMINAIRE RADIO PAR PAQUETS NOTE DE PRESENTATION VITA/PEP (PROJET D'ENTREPRISES PRIVEES)

N'DJAMENA TCHAD

R6daction: YONDAILAOU gue Tolloum, Administrateur du Projet Supervision: Mr. Iven L. OSE, Directeur du Projet Mme Luesette S. HOWELL, Directrice Adjointe du Projet

10 Juillet 1992

62 SEMINAIRE RADIO PAR PAQUETS

NOTE DE PRESENTATION

Cette note loin d'&tre exhaustive a pour objet de pr6senter 1'utilisation de la Radio par Paquets au Tchad, un pays en d6veloppement qui comme les autres pays d'Afrique connaft d.es difficultes enormes en mati&re de communication.

Notre experience.

Par l'action qu'il m6ne au Tchad ou it evolue dans un environnement souvent perturbe tant sur le plan politique (guerres civiles, instabilit6 r6p6t6es dot6es d'une ins6curit6, manifeste) et naturel (s6cheresse et famine), VITA/PEP (Projet d'Entreprises Privdes) avec l'appui de son siege a r6ussi une bonne op6ration en de communication en introduisant dans son programme la Radio par Paquets qui arrive A point nomm6.

VITA est la premiere agence d'ex6cution A apporter cette nouvelle technologie au Tchad. En tant que pionnier, it peut l'utiliser pour d6velopper ses propres activit6s en matiere de d6veloppement la ou it intervient, et 1A ou it peut intervenir, sans oublier l'aide qu'il peut apporter dans ce domaine A ses partenaires de d6veloppement pr6sents sur le terrain.

Lors de la mise en marche de la Radio, lesdits partenaires tant du secteur priv6 que du public ont 6t6 invit6s a assister a la s6ance de d6monstration. Its ont apprdci6 la nouvelle technologie, et suivant leur r6action, ils sont dispos6s & partager 1'exp6rience pour Eventuellement l'adapter A leurs propres besoins.

L'extension r6cente de la zone d'intervention de VITA/PEP A plus de 500 km dans le sud du pays avec 1 ' ouverture de son bureau r6gional de Moundou enrichira encore ses analyses, et d6veloppera ainsi ses chances de succ6s.

Le lancement de 1'opdration Radio par Paquets a commencd donc avec 1'extension des activit6s de VITA/PEP vers les provinces.

Plusieurs facteurs expliquent cette initiative:

- atteindre l'un des objectifs fix6s par VITA; & savoir le transfert de technologie;

- palier aux difficultds de communication inter urbaine;

- r6duire les couts d'utilisation des autres moyens de communication (T616phone, Fax, T61ex... ). I1 est A noter qu' un coup de tdldphone entre N'Djamdna et Moundou coOte 300 Fcfa soit $US 1.2 C par minute, et trois minutes de fax content. environ 7.000 Fcfa soit $US 28. Aussi, avec la Radio par Paquets si l'on rdussi A 1'utiliser rationnellement, on sera A mesure de r6duire sensiblement ces couts du moins de moitie sans parler des taxes qui sont in6vitables. 63 - rentabiliser les communications par radio (eviter les interferences au moment de l'utilisation de la voix);

- mieux coordonner les activites entre N'Djamena et Moundou;

- renforcer et faire une 6conomie des moyens de communication adaptables aux besoins des acteurs de d6veloppement a la base (ONG) op6rant dans les centres ruraux difficilement accessibles; - participer au renforcement du r6seau de communication classique et de l'infrastructure technique (radio et ordinateur) etc...

I1 est int6ressant de souligner que dans un pays comme le Tchad ou les infrastructures routi6res sont quasi inexistantes, et ou, le T616phone coOte cher, la Radio par Paquets peut ouvrir des perspectives int6ressantes pour assurer la communication au niveau de la gestion des petits projets ne disposant pas des moyens pour assurer la communication par T616phone et fax qui n6cessite des grands moyens et de 1'6nergie qui coute aussi chere comme 116lectricit6.

Avantaaes de l'utilisation de la Radio par Pactuets.

- Elle peut avec du matdriel simple et ad6quat etre facilement utilisee partout. M@me dans les endroits recul6s des zones qui ne disposent pas de 1'61ectricit6. Une batterie de voiture de 12 Volts peut faire 1'affaire; de qu'un ordinateur portable qui marche avec une batterie rechargeable avec les panneaux solaires.

- Elle est trios discrtte par rapport 8 la communication par la voix. En plus, it n'y a pas d'interf6rences avec les autres stations;

- Elle est accessible, quand l'utilisateur connaft d6j& l'usage d'un ordinateur ordinaire sans aucune sp6cification, it peut titre portatif m6me sans avoir le disque dur;

- Elle ne demande pas de grands moyens en matiere d'ordinateurs, on peut utiliser le mgme ordinateur qui sert 6galement 6 faire les autres travaux;

- Elle permet de transfdrer les informations et les donn6es par paquets. Cela dvite le retard A accuser pour 1'envoi du courrier par la poste; - Utilisation de la bolte 8 lettres en Pabsence du correspondant 8 qui on veut s'adresser;

- Bonne gestion du temps quand on donne des instructions au correspondent.

- Rdcup6ration des instructions sur 1'ecran si le message n'est pas bien requ;

64 - Imprimer lesdites instructions en meme temps qu'on les regoit sur 1'6cran;

- Envoyer les messages tout en donnant les instructions par la voix.

Dans le cas de figure de VITA/PEP, nous sommes a notre phase d'experimentation. Nos op6rateurs ne sont que des amateurs ayant r6ussi A manier 1'6quipement apr6s une formation sur le tas qui a 6t6 assur6e par le technicien qui 6tait venu installer le mat6riel. Ajout6e A cette formation, ils beneficient d'une assistance ponctuelle d'un consultant qui intervient lors de ses passages a N'Djamdna et Moundou ce qui n'a pas manqu6 de renforcer leur capacit6 de maniement.

Notre Eq_uipement et nos Fr6quences.

L'6quipement requis qui comporte une Radio TRANSWORLD, d'une TNC PSK-1T et tous les autres accessoires nous ont 6t6 installes par le consultant venu du si6ge. I1 s'agit de Monsieur David G. HENDERSON qui nous a 6galement guid6 dans nos premiers pas pour la mise en exploitation des appareils.

Nous avons dans un premier temps commenc6 avec comme frequences les 6.475.5 Kz et 6.970 Kz. Compte tenu des difficult6s rencontr6es en matiLre d' interfdrences et sur les recommandations du consultant, nous aeons laissd tomber la 6.475.5 et obtenu en remplacement la 9.475 Kz. Par cons6quent, nous transmettons et dmettons actuellement sur les 6.970 et 9.475 Kz.

Les r6sultats obtenus sont enti6rement satisfaisants dans les transmissions et par la voix et par paquets. Aux heures choisies; A 09 heures du matin et A 14 heures les apres midi, les .transmissions et Emissions se font sans interf6rences.

Difficultds/Probl6mes 8 rencontrer.

Conditions atmosphdriques (vents, chaleur, pluies, orages, temp#tes de sable...);

Coupures intempestives d'dlectricitd;

Taille des fichiers;

Interf6rences avec les autres stations.

Comment faire pour arriver A palier A ces difficultes et probli6mes.

I1 faut bien choisir ses heures de transmission;

Utiliser les UPS avec une bonne capacit6 de maintien en marche des appareils au moment de la transmission;

Limiter la taille des fichiers;

65 Se faire attribuer les bonnes frequences par rapport aux stations voisines-.

Conclusions et Perspectives d'avenir.

La mise en exploitation de la Radio par Paquets par VITA/PEP a rendu un tres grand service A beaucoup de ses partenaires et surtout aux acteurs de developpement presents sur le terrain qui rencontrent d'enormes difficult6s pour communiquer avec leur base a N'Djam6na.

On peut citer entre autres, CARE International, le Corps de la Paix, OXFAM, World Vision et l'Institut Chr6tienne de Linguistique. Outre ceux-ci, les organismes qui de pres ou de, loin ont vu et utilis6 ce service se sont interesses a cette nouvelle technologie et esp6rent 1'exploiter eux memes.

Notre bailleur principal l'USAID et l'Ambassade des USA ont tgalement b6ndfici6 du service au moment du deplacement de leur personnel dans le sud, tout comme le PNUD principal interesse pour avoir accept6 de financer l'op6ration.

Dans le cadre des perspectives d'avenir, VITA/PEP tAchera de passer de l'amateurisme au professionnalisme. Aussi, it sera question de recruter du personnel qui sera forme pour ne s'occuper que de la radio et A plein temps. Ceci permettra d'assurer la perennit6 de 1'experience qui a 6t6 initiee dune part, et d'autre part, d'intervenir auprbs de nos partenaires intdress6s par cette technologie.

Cette performance nous permettra Egalement A nous m8mes de mieux coordonner nos activitds qui wont s'agrandir avec les interventions prevues A partir du bureau r6gional pour d6servir toute la zone 6conomique du pays qui couvre les cinq (05) plus importantes pr6fectures du pays.

66 PACKET RADIO SEMINAR

PRESENTATION

The purpose of this presentation, which is far from exhaustive, is to describe the use of packet radio in Chad, a developing country which, like other African countries, is experiencing great difficulties in the communications area.

Our Experiment

Through its activities in Chad, working in an environment which is often disturbed, both on the political plane (repeated civil wars and instability, characterized by obvious insecurity) and on the natural plane (droughts and famine), VITA/PEP (Private Enterprise Project), supported by its central office, has succeeded in introducing a sound communications operation through its timely Packet Radio program.

VITA is the first implementing agency to bring this new technology to Chad. As a pioneer, it can use it to expand its own development activities, both in existing and in potential areas, not to mention the assistance which it may offer its local development partners in this field.

During the implementation of the Radio, these private and public sector partners were invited to attend the demonstration session. They appreciated the new technology and, based on their reaction, were prepared to share the experiment in order to adapt it eventually to their own needs.

The recent 500 km extension of VITA/PEP's area of activity in the south of the country as a result of opening a regional office in Moundou, will increase its analyses and promote its chances for success.

The launching of the Packet Radio operation thus started with the extension of VITA/PEP'S activities to the provinces.

This initiative is based on several factors:

achieving one of the objectives set by VITA, i.e. the transfer of technology;

solving long distance communication problems;

reducing the user costs of other means of communication (telephone, fax, telex, etc.). It should be noted that a telephone call between N'Djam6na and Moundou costs 300 CFA

67 francs, i.e. US$ 1.20 a minute, and three minutes of faxing costs approximately 7,000 CFA francs, i.e. US$ 28. If the Packet Radio is used properly, it would be possible to reduce these costs considerably, or at least by half, not including the unavoidable taxes;

making radio communications profitable (avoiding interference during voice transmissions);

improving the coordination of activities between N'Djamena and Moundou;

strengthening and making savings in the means of communication, adaptable to the needs of front-line development agents (NGO), working in rural centres with poor accessibility;

participating in the strengthening of the standard communication network and the technical infrastructure (radio and computer), etc.

It should be emphasized that, in a country like Chad, where the road infrastructure is almost nonexistent and where telephoning is expensive, Packet Radio may open up significant perspectives to ensure communication for the management of small projects lacking the means for telephone and.fax communication which requires substantial funds and high-cost energy, such as electricity.

Advantages of using Packet Radio

With simple and adequate equipment, it can be used easily everywhere, even in remote locations of areas lacking electricity. A 12 V car battery can do it, as well as a portable computer with a battery rechargeable by solar panels.

It is quite distinct from voice communication. In addition, there is no interference from other stations.

It is accessible, if the user already is familiar with a regular computer without any specifications; it can be portable even without a hard disk.

It does not require major computer equipment. A computer used for other purposes can also be used.

- It allows for the transfer of data and information in packets. This avoids delays in responding to items sent by post.

68 Using the mailbox if the correspondent to be addressed is absent.

Good time management when giving instructions to the correspondent.

Display of instructions on screen, if the message is not clear.

Printout of these instructions at the same time as they are received on screen.

Sending messages while giving voice instructions.

For a VITA/PEP profile, we are in an experimental phase. Our operators are only amateurs who have succeeded in handling the equipment after some training by the technician installing the hardware. In addition to this training, they also receive regular assistance from a consultant during his travels to N'Djam6na and Moundou which, certainly, has increased their computer skills.

Our Equipment and Frequencies

The required equipment, including a TRANSWORLD Radio, a TNC PSK- 1T, and all other accessories was installed by the consultant from the central office. This was Mr. David G. HENDERSON who also guided our first attempts when putting the equipment into operation.

We started with 6,475.5 kHz and 6,970 kHz frequencies. Considering the difficulties we encountered with interference, and on the recommendation of the consultant, we dropped the 6,475.5 and replaced it with 9,475 kHz. As a result, we are currently transmitting on 6,970 and 9,475 kHz.

The results are to our full satisfaction, both in voice and packet transmission. At the selected hours, 9:00 a.m. and 2:00 p.m., the transmissions are sent without interference.

Difficulties/Problems to be Solved

Weather conditions (winds, heat, rain, thunderstorms, sand

storms, etc.); -

Excessive power cuts;

Size of files;

Interference with other stations.

69 Ways of Solving these Difficulties and Problems

Transmission times must be carefully selected;

Using UPS that are fully capable of keeping the'equipment running during transmission;

Limiting the size of files;

Acquiring good frequencies in relation to adjacent stations.

Conclusions and Future Prospects

The installation of Packet Radio by VITA/PEP has provided a substantial service to many of its partners and, in particular, to local development agents who are facing considerable difficulties in communicating with their base at N'Djamena.

CARE International, the Peace Corps, OXFAM, World Vision, and the Christian Institute of Linguistics, among others, can be mentioned. In addition, organizations which, closely or from a distance, have observed and used this service, are becoming interested in the new technology and are hoping to use it themselves.

Our main sponsors, USAID and the U.S. Embassy, have also benefited from the service when transferring their staff in the south, as has the main UNDP involved by having accepted to fund the operation.

As regards future prospects, VITA/PEP will attempt to move from its amateur phase to a professional level. It will also have to hire staff who will be trained to work with the radio exclusively and on a full-time basis. This will make it possible to ensure the permanence of the initial experiment, and also to work with those of our partners who are interested in the technology.

This achievement will also allow us to improve the coordination of our own activities which will be expanded, through the planned activities of the regional office, to serve the entire economic zone of the country, covering its five most important districts.

70 Philippine Cooperative Rural Bank Foundation, Inc.

BOARD OF TRUSTEES 1990-91 OFFICERS z MR. NERNARDWO DIGNADICE, JR.

MR. LINO A Project Paper WO-Chaftm

MRS. NEMESIA SUAN on the

'MR RFGINOO.VERGARA MINDANAO PACKET-RADIO NETWORK ry FOR COOPERATIVE AND AGRIBUSINESS DEVELOPMENT ITTY.ISIDRO LICO DNetor

MR. ROMEO GARCfA b Ohclor ATTY. JOSEFTTO A. CUILLERMO DreaonAarp Eaeaim Oleo

Presented in NAIROBI, KENYA

: cR/ or /vlnaNON, rec. au or wlml corAUro, M. ceioriKaAmb omDfxrAL,D+G cR/ or w+AO oa NoRTt WG y cR/ or uZMD+c ca/ or rec. i CR/ Or AIMN, DIG G

CR/OINORTMCOTA/AT41.W_

: - CR/ 0/ M1tAIHlt ORDMAl.INf CR/O/AL/AY,D1C. CR/O/DAIW,RdG r Q1/ or /01101. Mr CR/ O/ DAC06 NORTC ING CR/01DAVAOCRY,LVC. Q/ DCL KX. WG CR/ Or DAVAO. DIG AGUSAN 10ORTIC-11MAN CRY C7R Nr- Presented by r

Atty. Josetito A. Guillermo Executive Officer Philippine Cooperative Banks Foundation, Inc.

August 1992 x c. 71 Philippine Cooperative Rural Bank Foundation, Inc.

BOARD OF TRUSTEES 1990-91 OFFICERS

MR. BERNARDINO DIGNADICE. JR. ChMmUn LIST OF ACRONYMS MR. LIMO R.?E1VAFWR

MRS NEMESIA SUAN Ttwurr 1. PCRBFI - Philippine Cooperative Rural Bank Foundation. Inc. MR.REGINO O. VERGARA 2. ACDI - Agricultural Cooperative Development it .ISIDRO LICO Dlnelor International

MR. ROMEO GARCIA Oweew 3. VITA - Volunteers in Technical Assistance

ATTY.JOSEFITO A. GVILLERMO DYteb7AcOV Emaffm ORar 4. VOCA - Volunteers in Overseas Cooperative Assistance

5. DAFKDACO - Davao Federation of Agricultural MEMBER BANKS Cooperatives. Inc.

6. PLDT - Philippine Long Distance Telephone cite of sUKIV%W% NC Company CRR Or SOU111 COTARATO. NC T. NCR - National Capital Region (Philippines) CRR or UnAmn OOCWVff AL. NC CRR DEL NORTE NC a. GOs - Government Organizations CRN Or NECROS oCODENTAI. IVC CRR OrLrrM NC 9. NGOa - Non-Governmental Organizations CRS OF CAMICLI%. 1%r. 10. MFCB - Mindanao Federation of Cooperative CRS Or ARL X NC Banks. Ltd. CAR SEMMU CRS. Nr. CRS OF NORTH COTARATt4 ryC. 11. MAN - Main Area Network CRS orM sAMb ORtFK AL. r4 Network CRR Or ALRAY. NC. 12. LAN - Local Area CRR Or RMLO. NC. 13. PBSP - Philippine Business for Social Progress Cu Or S011OU NC CRR Or RAM %ORTI: K 14. CIDA - Canadian International Development CRS OF DAVAO MY. r4C_ Agency

CRR OF DAVAO DEL KT. NG

CRR Of DAVAO. NC. ACUSAN NORT&KIMAN CITY C796 NC.

72 BRIEF PROFILE OF THE. PHILIPPINES:

The Philippines has approximately 7,100 or more islands with an estimated population of about 60.5 million as of 1990.

It is divided into three maJor islands namely: Luzon,

Visayas and Mindanao (please see map).

PROFILE OF MINDANAO:

Mindanao has a total area of 94,630 square kilometers. Area wise it is bigger than Singapore's 581 square kilometers,

Brunei's 5,800 square kilometers, almost triple of Taiwan's

36,000 square kilometers, larger than Belgium's 43,075 square kilometers.

There are 23 provinces in the Mindanao Region listed as

follows: Agusan del Norte, Agusan del Sur, Basilan, Bukidnon,

Camiguin, Lanao del Norte. Lanao del Sur, Maguindanao, Misamis

Occidental, Misamis Oriental, North Cotabato, South Cotabato,

Sultan Rudarat. Sulu. Surigao del Norte, Surigao del Sur, Tawi-

Tawi. Zamboanga del Sur. Zamboanga del Norte and Sarangani.

Of the 23 provinces, there are four island provinces:

Camiguin. Basilan. Sulu and Tawi-Tawi. There are sixteen (16)

cities namely: Butuan, Cagayan de Oro, Cotabato, Dapitan, Davao,

Dipolog. General Santos, Gingoog, Iligan, Marawi, Pagadian,

Oroquieta, Ozamiz. Surigao, Tangub and Zamboanga. Of the 16

cities. Davao City is the biggest in area, and considered as the

third developed city in the country.

Mindanao has 3.73 million hectares of land, make-up 38% of

the total farm area in the Philippines. Farmland per capita in

Mindanao is 2.624 sqm. per person as of 1990. which is sixty-

eight (68x) more than the rest of the country, excluding the

73 REPUBLIC OF THE PHILIPPINES Political map

rAr.r. "'7N

INDIAN OCEAN

., fIRJIAN !/A

t+.Krns

Yi1MAr Ml M'.

74 In»s Y m»»C2p7>t (>1 C)m m O Zi>>i 4 z ai O o O m Z Z= > Z s b i z g s z z RQ m n m m O a ^D re$ C m m © I ME] o

Q y Y 0 y 0 (1 r + S= b rM y1 C a' YY $4 s L b C 3 y iii f1 f1 l S b T V h n<

H 2 N f S r

O O H y N r v ! r y P r+ Y M M M Y M P M N y+ P Y P Y Y r r V O M N Y N N U O a M r M N O N O N W r Y N N r P N Y b N P M 0 0 Y r 0 N v N v y v P O Y r Y r y r y O V N r O V+ p P O P Y N A/ Vt V Y s u p 1 0 0 O P O fJ O N U O N ip t!U r P ! fi (J T

N _ N r V O O N O O Y 41 V I. T m 01 N O. O P N

r r.. r N r r 40 . y P. N r N r r r.. . r . . r . . m y N . . r r . y N . ..

S Y O V b °1 O+ O O N 41 V Qj t P g a t + =' a r i V N P u + P v+ V /J b 6 P v ti Y IJ O r P + O r q qL,'' P O N b N V y N P N V O m V V

C 2 m f gg 2- S ao m_ i r r r r e p r '=--x-Q m x Q m> r , i- i i i p- i r $o xooo ! s r b r r O a r g g o rr Y i 3 m g g$

n a r r f. g > > o m National Capital Region (NCR-Luzon). Fifty-five percent (55%) of unutilized lands are reported to be in Mindanao.

Mindanao has a large agricultural base, producing about 40% of the agricultural output of the entire country. Mindanao's forests contributed 73.4% of the national value added in agricultural crops. Mindanao's Regional Product is 45.6% agriculture, 21.7 industry, and 32.5% Services.

Export receipts of Mindanao from January to September 1991 was US$701,096,393.00 FOB, 10.9% of the total national export receipts. National Capital Region (in Luzon) accounts for about 65% of the total export.

Inspite of these figures, it is sad to note that maJority of the people in Mindanao live below the poverty line. This mainfests that Mindanao's vast resources is unevenly distributed.

MINDANAO'a COOPERATIVE MOVEMENT

Being aware of such a sad plight, the Mindanaoane, out of their initiative and with meager resources, bonded themselves to form various cooperatives to empower themselves and the rest of the less priviledged. As a result, Mindanao's cooperatives are leading in the country's cooperative movement and still growing to offer various services, deliver peoples' basic needs, hence offering effective alternatives in alleviating poverty and fulfilling the peoples aspirations for self-reliance, free from the clutches of unscrupulous big business whose main reason is to merely generate profit.

Leading the Mindanao regions' 23 provinces, and playing the maJor roles of cooperativism are: Davao del Norte, Davao del

Sur, and South Cotabato. Their successes are being shared through replication of appropriate technology, exchanges of

76 business ideas, continuing hands-on training programs, and the like.

These Mindanao cooperatives offer distinct and varied

services. They acknowledge that varying ' degrees of

interdependence from each type of the following cooperatives

operating within this maJor island can foster mutual relations

and interest for a common good. These cooperative types

comprising the movement in Mindanao are: agricultural, coop

banking, credit, aqua-culture, marketing, electric, hospital,

telephone, transport, employees' coop, co p.m d Most recent VA CUur cooperatives developed pioneering n in the country are, cooperative air, cooperative hospital and cooperative media.

These cooperatives are geographically dispersed. They are

situated in both rural and urban areas. Coordination is

difficult among the coops, non-government organizations (NGOs) as

well as government agencies. The absence of an effective communication system slows down the desired development, hampers the flow of much needed business as well as urgent information.

The Philippine Long Distance Telephone Company (PLDT), the only company in the Philippines engage in the installation of telephone system does not really reach in more provinces. If they happen to install a telephone system, expect a very poor service. It does not help at all.

Communicating with the coops in the provinces is through the postal service which would take 4-5 days delivery and depending on the choiced location. cable gram is available. The most realistic is personal visits which could be very tiring, time consuming and costly. Cooperatives in the Davao Provinces were in the.midat of this problem, more so with the remote provinces in the island.

77 ORTCAT. BACKGROUND OF THE. COMMUNICATION SYBIL

In 1989, in order to solve the communication system' problem, series of consultations were conducted to gather feasible ideas \ the Davao Federation of Agricultural 1%441 Cooperatives, Inc. (DAFEDACO). This cooperative federation +t3#U'r

28 big cooperative members located in different parts of Davao

Province. The Cooperative Bank of Davao City has been an active

member of DAFEDACO.

In one particular meeting attended by Board of Directors,

Officers and members of the DAFEDACO, they come up with a

resolution to set-up an effective communication system . to

facilitate the smooth flow of information among the coops, NGOs

and GOs, and for faster exchanges of much needed agribusiness

data.

The Voice Communication System using a VHF/FM Repeater had

to be installed on top of the highest Mawab mountain in Davao del

Norte. With this it could already contact to almost all of the

member coops and other/\aagribusiness related entities in the Davao

Provinces.

However, these communications equipment were very expensive.

Hence. DAFEDACO formulated a project proposal and had it

submitted to the Philippine Business for Social Progress (PBSP), the a non-governmental organization' to source funding for

project.

Later that same year. PBSP was able to source the needed

funds from the Canadian International Development Agency (CIDA).

DAFEDACO was able to generate USS8.5 thousand in assistance

from CIDA. It was used to procure the following: power. 1) One (1) unit Repeater VHF/FM 80 TO 100 watts V-100 95 watts 2) Two (2) units Icom Base Transceiver model power

78 DAFE D/\CO REPEATER MAIN AREA NETWORK:

1 - DUPLEX MODE CHAN 2 - SIMPLEX MODE

DO RAM CO

LEGF-ND:

BjA5E I COM V-100 - ft5 WATTS POWER OUTPu7

HANDSET I C01\4 H-16 - S WATTS POWER OUTPUT WITH EXTERNAL ANTENNA; 3.$ DB GAIN -

79 DATE DAC O LOCAL AREA NETWORK:

LEGEND:

BASE I COM V-100 45 WATTS POWER OUTPUT

HAND5ET5 ICOM 14-16 5 POWER OUTPUT W/ RUBBER DUCKY ANTENNA

MOBILE. ICOM H-16 5 WA-TT5 W1 SO WATTS BOOSTER OR EASE I COM V-100 Stn 3) Seven. (7) units Portable Handsets Ieom model H-16 V 5 watts power

4) Two (2) units Telescopic Towers 80 Ft.

5) Two (2) units Vari-Loops Antenna; 9.0 db gain; Omni- Directional

With these communications equipment, the present communication system was set-up as follows:

A) Main Area Network (Please see diagram illustration.)

B) Local Area Network (Please see diagram illustration.)

There are three frequencies being used as follows: Two for the Repeater; one for Receive (RX) and one for Transmit (TX) called Duplex. And one frequency for the Simplex mode. All these frequencies are operating in the 160-165 Mhz commercial bands. In this system we use three channels as follow:

A) CHANNEL 1 DUPLEX PLUS

B) CHANNEL 2 DUPLEX MINUS

C) CHANNEL 3 SIMPLEX

A) CHANNEL 1 - It is used to receive and transmit via a VHF/FM

transceiver thru the Repeater in Mawab mountains

in Davao del Norte on any particular stations in the Main Area Network.

B) CHANNEL 2 - This is used primarily for the Local Area

Network, to contact personnels in a particular

station. It is actually the reverse of Channel

I without using the Repeater.

C) CHANNEL 3 - This is usually used when the Repeater lines are

busy, the stations using this channel must have

a base transceiver usually 45 watts. This is

called the Simplex mode.

81 PHASE IL N AREA NETWORK: PHASE Z

.N 1 DUPLEX 2. I PLEC rl S M MATI MAt,AY5ALAY DAVA0 OR. OUKIDNON

L. U PON I cAMtCkLAIN DAVAO OR.

'T&G U M C,&a DVO. KORZE AYA N i 1 Du 1 VAN Ro cIT`f Ar-LOAN NOR-Tf

N00

PCR.BPI TAG,t R ON6j sU L.TA M O Rbgt ltETA DAVA O C I T `t' K V DAP-AT ` v[3cD LArNA4 N CgLT6 M 15A W S occ

K I DA PAWAN pAVQO CITE' WoQTR WRA

LEG EN D: CALINAN MALI-TA PWAO CITY t>iwAo 5uR REPEAT Ei?

D I c-ro$ DL&vAO SL4 R HANDSET5 PCR,BFI LOCAL AREA NETWORK

CHAN 1 DUPLEX CHAN 2 SIMPLEX

CB DC MO6lLE pASE.

LEGEND:

BASE. CENTRAL

BRANCH OFFICE

HANDSET

FLOW OF MESSAGES

83 PHASEI PHASE I[ MT. PAG IYASGAS, DAVAO DE L SuR MT KITANGLAD, OUKIDNON DA F EDACO RPT REPEATER REPEAT ER c9 Ld BukIDNON CANIGI.IIN

c0 QtAT LAX" CITY

PC RORI DAVAO c 1 TY

CALliJW Dvo.clTy LEGEND:

R PT REPEATER

BASE STATIONS

PORTABLES LINKAGEOF PCRBFI COMMUNICATION 5Y5TEM .4 FLOW OF MESSAGES DAFEDACO NETWORK ___. DIRECT LINE NOTE.: P"ASE 2 RePEATER ALREADY INSTALLED. THE SYSTEM INSTALLED THRO ACDI. VITA-' & VOCA"s ASSISTANCE-

The communication system installed at the Davao Province triggered the Philippine Cooperative Rural Banks Foundation, Inc.

(PCRBF) and the Mindanao Federation of Cooperative Banks, Ltd.

(MFCB) to set up a communication system for the Cooperative Banks not only for voice communication as well as for Packet Radio and Phone Patch.

The pro,lect aims to enhance the flow of commercial activities, create linkages and/or promote business consolidation among the cooperative rural banks and the allied cooperatives and organizations.

The following steps were resolved for further expansion and

improvement of the communication system:

1. Set-up a communication system not only in a particular

province but also for the entire Mindanao islands. This

was be broken down into two (2) phases:

A) Phase I - Installation of a Repeater in Mt. Pagiasgas,

Davao del Sur, to cover the southern part of

Northern Mindanao (finished)

B) Phase II- Installation of an additional Repeater in

Mt. Eitanalad, Bukidnon, to cover the entire

Mindanao Region (pls. see diagram illustration).

2. Link the Communication System to a Terminal Node Controller

and interfaced with the computer and phone patch (See illustration).

3. Link some of the stations in the DAFEDACO Network (See

illustration).

85 In June, 1990, PCRBFI prepared a project proposal and submitted to ACDI for funding. In November 1990 ACDI was able to extend a grant assistance package with corresponding counterpart from PCRBFI. VITA extended part of the technical assistance package through ACDI.

The following communications equipment were procured thru the said grant:

1. One (1) Unit Renwood Repeater between 80-100 watts power

2. Two (2) Units Antenna System 9.0 db gain Omni-directional

3. Two (2) Units Telescopic Towers with 60 & 40 Feet. 4. Two (2) Units Terminal Node Controllers

5. One (1) Unit Power Supply; 20 Amperes; 13.8 Volts DC The first of the project, the installation of the Repeater in Ht. Pagiyasgae. Davao del Sur is already finished and some of the networks bases. ' There are three frequencies used and three (3) channels. Channel 1 is for the Repeater System; Channel 2 is the reverse of Channel I and Channel 3 is the Simplex mode for direct communication among its base stations.

BUSINESS APPLICATIONS, STRATEGIES AND ACTION PLAN FOR THE COMMUNICATION NETWORK:

As envisioned, the following business applications were perceived to be enhanced through the use of the communication system:

1. Banking Applications: With the establishment of a regional- level cooperative development bank serving as many as 13 provinces with separate branch offices in each location, on-line computer capability will be a requisite to efficient daily banking activity. A standardized management information and reporting systee shall be developed which serves the routine business functions of a proposed regional cooperative bank

86 network. This includes the necessity of developing a

communication network with computerized functions for savings and

loans accounts, general ledger activity, cost accounting,

payroll, etc. It shall also extend services for fund transfer,

financial clearing and interlending. Standardized software

programs designed for financial institution application are

available but need to be customized to serve the particular needs

of the coop bank/s in Mindanao.

2. Cooperative Marketing Information System: The voice/data communication network has the hardware capacity to collect and disseminate critical data and information necessary assuring cooperative enterprise a competitive position in the marketplace.

This includes inter-cooperative tradl and trade between cooperatives and non-cooperative businesses. Commodity demand and supply data and wholesale and retail price information is particularly important for agricultural commodities produced, processed and marketed by cooperatives.

3. Information Exchange and Communication Applications:

Additional business applications offered by the presence of voice/data communication network extend to such areas as technology transfer, problem solving, data transmission, oral and written communication, personnel supervision, program management, education and training, and emergency information transmission.

Such applications also facilitate enchanced cooperation and understanding between user groups, thus contributing to stronger business and institutional relationships.

Stages in Communication Network Development:

1. Equipment Hardware Installation and Testing

2. Business Application Software Design, Installation and Testing, and

3. Personnel Recruitment, Training and Performance Evaluation.

87 FUTURE PROSPECT FOR IMPROVEMENT:

With the trend in state of the art Satellite Communication

System, linking the current communications network with the satellite can be a reality in the near future (please see illustration). Granting that fund and technical assistance from various sources would be available, Mindanao known for being the

Land of Promise shall emerge and make this dream a reality.

CONCLUSION:

As U.S. handedly and astonishingly defeated Iraq mainly because sophisticated communication edge in Desert Storm,

Mindanao will certainly WIN the 'WAR against Poverty in Communication.

Respectfully submitted:

ATTY. JOSEFITO A. GUILLERMO Executive Officer Philippine Cooperative Banks Foundation, Inc.

88 TRANSVERT£R

MlcaoPho"M

BASE RADIO El Is t TMANSI VER POWER SPI CR, SvPpl-Y t-sI w KEY a0ARD

PCRBFI'S FVTuRE COMMUNICATION SYSTEM -TO bE SET UP AF IN1 AP OUT AF OVT AF IN -TEI-EPMONS PAC COM AUTO DTMF PACKeT NOTE.,. 5A'TI±LLI'tE OISG AND AVTO PItONC- PATCH PATCH MODEM /.R6 TAP ones NOT Y6't INSTALLep.

AS IN - ANDIO FREQUENCY IN AF bu't'- p.U010 FIkF-QVENCY OUT PACKET AND SATELLITE RADIO IN SIERRA LEONE

INTERNATIONAL WORKSHOP ON DIGITAL

RADIO TECHNOLOGY AND APPLICATION:-24 - 26 AUGUST 1992

INTRODUCTION

Sierra Leone is a small country situated along the West Coast of

Africa as you can see from the world map. Its capital is Freetown.

Let me tell you a little bit about PLAN International where I am working as Management Information System (MIS)Manager and in the

Accounts Department. I am in charge of the satellite operation in uploading and downloading files from our International Headquater,

Rhode Island U.S.A and the Regional Office at Dakar, Senegal via

VITA, Virginia, USA.

PLAN is a non-profit, non-religious and non-political international development agency that seeks to help needy children and their families by improving the community in which they leave. The goal of the organization is to provide needy children with the necessary security, dignity and opportunity to become healthy and active adults who will contribute to the development of their society.

In general, PLAN proaK,tes small-scales, people-participatory project activities. Each of the program locations are staffed by locally hired employeev who work closely with the Foster Families and their communities. The staff provides training, motivation, and guidance to the communities and encourages people's involvement 90 in all stages of planning, implementation and evaluation.

PLAN International began its activities in Makeni Northern

Province, Sierra Leone in 1977 under a government agreement with the Ministry of Rural Development, Social Welfare, and Youth. A second program was opened in Freetown the capital city in 1980. A third program in Moyamba Southern Province in 1989 and the fourth was opened in Freetown for the urban poor areas (slums)in 1991.

PACSAT GROUND NATION FREETOWN BACKGROUND:

Let's come now to the installation of the PACSAT Communication

Experiment Ground Station. In early 1991, PLAN International

Management Information System (MIS) Regional Coordinators for East and West Africa, together with Eric Rosenberg from VITA PCE (PACSAT Communication Experiment) GROUND STATION, Arlington, Virginia,

U.S.A. came to Freetown in PLAN International Freetown Field Office to install the. :system.

LICENSING FOR _ TLLE_ f1BQU D _:iTAT1QN

PLAN International Freetown was licensed on uplink frequencies in the 148 KIZ band with downlink channels in the 428-430 MHZ band . Theve frequencies we!re authorised for LEO satellite used by the International Telecommunication Union (ITU) in Geneva, Switzerland.

91 Upon authorization and licensing from the National

Telecommunications Commission on the desired frequencies, and based upon a report of "clear frequencies" in the 428-430 MHZ band,

installation of this system was completed at PLAN"s offices in

July, 1991. During this process, a carrier was heard at 430 MHz, with considerable sidebands above and below the frequency. The

carrier was noted to be continuously operating.

INTERFERENCE:

At the time of installation the transmissions were possible only in

amateur frequencies and not VITA frequencies. There was severe

static interference. We tried earthing; grounding the various

equipment however succeeded only in marginally reducing the static

interference. Then Eric left for USA. There seemed to be a

software failure and we were not able to have a single successful pass after Eric left.

Due to this interference, the Ground Station was unable to

communicate with the satellite. VITA/PLAN then scheduled a second

trip which was to determine the source of this interference. Then

came the second person from VITA. USA, Nark Oppenheim. The source

was tracked to a site about 10 miles from the Lungi International

AirF,ort Freetown-Sierra Lcione. This was determined to be a horizontally polarized signal, centered at 430.0017 MHZ, within the internationally allocated Amateur Radio band.

92 Then came Marie-Ange our Regional Coordinator for West Africa in

March 11-20 1992. She examined the ground station and it showed that Keplerian Elements had not been updated since October and that two wires connecting the antenna to the rotor were loose. In the process of re-wiring the rotor the Kansas City board was damaged, due to the fact that the cables were having the same color of paint. Upon replacement of the board and the re-calibration of the antenna the ground station worked perfectly.

The problems of the ground station rotor as mentioned earlier had two loose wires. These wires most probably got loose due to the cables getting stuck during rotation and thus pulling the wires loorse. In order to further minimize the possibilities of the antenna cables getting stuck on the tower, the sharp corners of the tower held been (sawn off, putty was used to fill the nooks on the corners of the tower and to further round off nails and screws-

It was noticed that during testing of the rotor controller there were attempts to force the rotor controller in the target position whenever the antenna stal led. The Regional Coordinator warned that Whe:tiever the antennb stalls no forcing should occur, the operator should go up on the roof to make sure all the cables are free and clear. This will minimize wear and tear of the equipment as well as the need to calibrate the antenna positions.

93 After Marie-Ange rectified the fault, the PACSAT Communication is now working well. Messages can now be uploaded and downloaded freely without any problem. I can send files to our International

Headquarter (IH)in the US and the Regional Office at Dakar, Senegal etc. via VITA as quickly as possible. I think if we have a satellite station at our International Headquarter and the Regional

Office, work will be more quicker than usual. The PACSAT

Communication Experiment is very good and has indeed eradicated the complexities associated with International Communications/Data transfer.

The PACKET Radio is a method been used in sending messages and any file on a computer diskette via radio with other Field Offices that are about 200 or more miles away from our office in the city- The computer runs a special software program that allows communication with the modem and the other stations.This can be done in few seconds and get response at the same time.

UQ,,c.RST T ONS FOR

1. Remote(cordless) control of the antenna, so that the only

cable connected to the antenna is the power cable. Provided

this does not reduce the cost - effectiveness of the project.

2. Better and smaller type antennas, so that the ground station

operator can automatically "View" the antenna in motion to

determine whether they STALL or not.

94 3. Improve compactness of the ground station equipment, so that

installation/re-installation work and time may be considerably

reduced.

4. For the ground station software: To use internal parameters to

tag " obsolete" files and delete them without operator's.

intervention. This will save operator time and prevent

program error if the ground station software is downloading

files too big to fit in the remaining disk space.

5. Reduce radio noise during transmission to avoid distraction.

Submitted by: Joseph Sandi - MIS Manager.`

PLAN INTERNATIONAL FREIiTOWN,SIERRA LEONE

95 HOW A PACKET RADIO COMMUNICATIONS SYSTEM HAS SERVED ISERST

ResearcherDJAMA DIRIEH ISERST DjibOUti July 11, 1992

96 A. INTRODUCTION.

A.I. DJIBOUTI The lie public of Djibouti is one of the world's 42 least developed countries (LDCs), as classified by the United Nations. It is situated on the East coast of Africa between Ethiopia, Eritria, and Somalia at the mouth of the Red Sea and the Gulf of Aden at a latitude of between 11 and 12 degrees NORTH, with a surface of 23,200km square and 370km of coastline.

Climate Djibouti has a tropical climate, and is officially classified as a maritime desert; a climatic region that is hot and humid in the winter and extremely hot and humid in the summer. Rainfall Annual rainfall in Djibouti is not very high. It varies between 50mm and 300mm per year.

Population Djibouti's population was estimated to be 5100000 in 1991, and is projected to increase to 700,000 by the year 2004 (at the current yopulation growth rate of 3.0%). This growth rate has decreased rom that during the 1970s and 1980s when it was at 5.6%.

Natural Resources. Although Djibouti has no known resources of Qetroleum that are commercial viable, it has a wide range of indigenous sources of energy that could be profitably exploited in an environmentally sound manner. Djibouti has good solar energy, wind energy and excellent geothermal potential. The known natural resources are limited and they consist essentially of salt, calcium perlite, diatomite, gypsum as well as minerals which could be useYur for the production of cement.

97 Djibouti's Communication Systems Djibouti is a focal point of communications commercially exchanges between Europe, Asia, in the cultural and Red Sea and the Indian Ocean. and Africa through the Telecommunications. The Republic of Djibouti has reliable and efficient telecommunications systems, which are managed by the Office for Post and Telecommunications (Office des Poste et Telecommunication OPT The OPT uses telephone,telex, telefax, exc and telegrapA angel. Although Djibouti telecommunications systems has a relatively large international activity and a low level of internal activity, each regional head uar ers has a postal and telecommunications agency linked o the Postmaster's Office. . International telecommunications traffic is managed through the Societe des Telecommunications Internationals de Djibouti (-STID) a subsidiary owned jointly.with a French two centers. Cable Company. STIR has i- A satellite telecommunications center situated at Ambouli linked to the INTELSAT and ARABSAT networks. This center has been financed jointly by Saudi Arabia and France from 1979, and utilizes the latest telecommunications technology. This has enabled Djibouti to link-u to a variety of so histicated international telecommunications networks, mainly to the Arab and Western worlds. 2- The International Maritime Transit Center which is situated in the city center. The center in linked to the SEA-ME-WE (South- East-Asia-Middle-East-Welter-Europe) undersea cable from Singapore to Djiibouti and from Djibouti to Marseille (France) through Jeddah ($audi Arabia) and Alexandria (Egyypt). This roject was realized with Saudi Arabian financial credits in 19M1. This has enabled Djibouti to link-up to Western Europe,the Arab region and the Pacific basin countries. It has also expanded and improved the services of the existing coastal line stations, as well as the national and regional communications of the East African region.

1. ISERST. 1SERST (National Institute for Higher Scientific and Technology Research) Djibouti East Africa, was established by presidential act in 1978. ISERSY is charged with coordination of all scientific and technical research in the county . It's role is to encourage and assist applied research in areas hat will improve the economy of the country as well as the standard of living of the Djiboutian eople. Also it's role is to introduce new technologies and promote heir use.

98 ISERST is a multi-disciplinary institute consisting of four different departments ( see attached flow chart) comprising; 1- k:arth Sciences 2- Human Sciences. 3- Natural Sciences 4- Renewable Energy and Technology. ISERST has been working with various aspects of developing the country's indigenous resources energ solar energy wind power and geothermal energy, since 1980 in co aboration w h a variety of international organizations. ISERST has tested and installed solar pumps, lighting systems and refrigerators. ISERST has also recorded an average wind speed up of 4m/s in most of the country. Such wind speeds are more than sufficient to power windmill pumps. The results of the scientific studies conducted by ISERST and AOUATER, and the exploratory phase involved in the drilling of six boreholes to a depth of 15O U to 2000 meters, conducted by ITALY and France have proven the potential commercial viability energy. of geothermal

2. ISERST Experience with Electronic Devices.

2.1 Existing electronic devices. Mindful of the computer's capacity for varied office applications, the management of the ISERST has equipped each section with stand- alone microcomputers. These stand-alone microcomputers are set up to serve and increase the productivity of the sections. The size and the type of the program application of these electronic devices depend upon the objectives which each section is expected to achieve. The software packages that run on the stand- alone micros include; 1- Word processing and desktop publishing. 2 Database management and integrated systems. 3- Graphic packages. 4- Communications systems. Thus the purpose is to provide for their own sections: 1- Wordprocessing, text editing and filing data capabilities.

99 2- Financial database management capabilities 3- Satellite communications capabilities. 4- Graphic capabilities and the use of geographic systems. 'information 5- Bibliographic database capabilities. Following a request made by ISERST, VITA installed a Geographic Information System (GIS) at ISERST. The ISERST GIS called At3as*GIS installed on a PC. lab has software The ISERST staff of the GIS laboratory have given GIS presentations to various ministries of the Djibouti government. Four services are most interested in GIG appp ication: Electric company (EDD Ministry of 'Public Work (TPJ Postal and Telephone Service (OP and water Distribution Authhority (ONED). Also the Ministryy o Health and Hygiene has expressed it's interest. All the Love services have to store, update, retrieve and analyze carto%raphically related tabular data information. ISERST staff will be able to assist these services with training and data base creation, once the services purchase their GIS software & hardware. Also a member of the ISERST staff made a presentation to the ISERST Management of an office based Electronic-Mail connecting the management of ISERST to VITA Headquarters. The management it's interest. has shown It has been recently agreed with the management of the Djibouti OPT to make a short demonstration of data transmission and reception between the Djibouti OPT satellite office to the VITA Headquarters Arlington V rginia. This has been tentatively arranged for 11tH August, 1492.

3. ISERST Packet Radio Communication Service.

3.1 ISERST/VITASAT GROUND STATION. ISERST sees itself as a catalyst in Djibouti for the introduction of new technologies in the country in this regard it was one of the first institutions in East Africa to install a packet-radio ground station for communications linked to a low-earth orbiting satellite. On 3rd December, 1991 the Djibouti government authorized ISERST to install and operate, on an experimental basis, a VITASAT ground station for the purpose of communicating with the UoSAT-3 Low-Earth orbiting satellite.

100 With the help of the VITA Communications Specialist Mr Eric Rosenberg, the technology section of ISERST installed the VITASAT ground station successfully on 12 December 1991. This project has enabled ISERST to linkup with the other operational VITASAT and Satellife stations in Kenya, Mozambique, Pakistan, Tanzania, Somalia, Sandia lab, Vita, ...etc. ISERST's principal interest for the introduction of the packet- radio communication service is to facilitate and to assist the Governmental and non-Governmental organizations with the exchange of technical information relating to health, energy, education, environment, and other humanitarian activities where the communications systems are inadequate. The Intergovernmental Authority on Drought and Development (IGADD), which consists of six (6) ast African countries, including Djibouti, Ethiopia Kenya, Somalia Sudan, and Uganda, has it's a ministrative eadquarters located here in Djibouti. IGADD also houses the Early Warning for Food Security Project (EWFIS). They are highly interested in exploiting the ISERST/VITASAT ground station communication services. IGADD has established a comprehensive sub-regional information system that facilitates the collection, analysis and distribution o information on natural resource management, environmental protection and desertification control, (IGADD-INFONET). This network has technical nodes in the IGADD member states called National Information Nodes (NINs) and it is necessary to transmit data between them and the IGADD headquarters. These are mainly binary data. As conventional transmission lines between the IGADD member countries are unreliable, both IGADD-INFONET and EWFIS have asked ISERST and VITA to carry out a demonstration/test of data transmission and reception between two IGADD member countries through the ISERST Packet-Radio communication system. IGADD also has requested these two partners (IGADD/Djibouti and IGADD/Kenya) to re-upload the binary data files they received from ISERST back to IGADD/Dj ibouti via satellite in order to evaluate reliability of the ISERST/VITASAT Packet Radio communication system, and the safety of the original files. On behalf of IGADD, the ISERST Packet-Radio communication system has enabled ISERST and VITA to cart out successfully a series o data transfer (53,924 and 160,000 bytes) between IGADD/Djibouti to IGADD/Kenya and to VITA Headquarters, on 10th May,1992. As part of the test/demonstration, we (ISERST) have recorded and re-distributed to IGADD/Djibouti and VITA the following: 1- Date and time we (ISERST) uploaded each file.

101 2- Number of passes it took station. to upload each file from ISERST 3-- bate and time we re-downloaded each file satellite to ISERST station. back from the problems we encountered uploading 4- Any and re-downloading the S- Date and time the ((VITA and KENYA ) received the first component of theile (or the complete file one pass). if received in 6- Number of asses it took to download each file to their stations. from Uo-SAT-3 7- Date and time they re-uploaded the files stations. from their respective 8- Anyy blems they encountered eac? Me. downloading and re-uploading Based on the success of transmission the first test tdemonstration of the data exercises phase 1 at ISERST, IGADD is considering the possibility of implementin a VITASAT communication system of the six -IGADD member states and for international connection.in each In this case, the antenna must be movable for tracking the satellite. Trackin of the antenna as well as automatic communication to the satelli a will be controlled by the station computer.

4. Technical and Human Resource Challenges

4.1 Technical Challenges The many technical challenges we meet in this field include: 3. The ISERST ground station is licensed to operate with the nominal u plink frequency of 148.mha and with the nominal downlink frequenc of 429.mhz from the satellite. The downlink varies +/- Skha from the nominal frequency moves from horizon to horizon. In this regard as theis imperativesatellite that the entire 16khz downlink bandpass it be clear interfering signals. of During the setup and testing of the station by Eric Rosenberg, a VITA satellite communications specialist, a signal was found to be interfering with the satellite's downlink transmissions.

102 The offending transmissions are short bursts of data sent on a regular basis by one station, often with a response from a second. effectively block UoSAT-31s transmissionThese transmissions for almost half the time it is in our view. It appears that the signal comes from a French military base about 2km south East of ISERST. Thus far no remedy has been found for this problem. 2. The six (6) month license we received from the Djibouti OPT which limits our use of the ISERST ground station to "scientific" communication onlyy, has decreased greatly the quantity of communication traffic through the use o the ground station.

4.3 Human Re-sources Challenges. Among the many challenges we face in this area are; 1. The lack of adequate rofessionally trained personnel of all specializationo rely ad to computer network management, particularly hardware and software maintenance. 2. The lack of adequate facilities and personnel for operating separate stand-alone com uters. We have experienced a number of breakdowns due to Lime sharing equipment misuses by unqualified people. 3. The lack of personnel and facilities at other national institutions and departments with which the ISERST would be corresponding and exchanging information in it's quest for centralizing information at the national level. Examples are t-he GIs and Packet-Radio Technology Applications s stems that have not really taken off for lack of inputs from the other Government departments. The challenge will be the mobilizing the concerned de artments in the development of a national inter-active information network in these fields. For this, eventually professional personnel will need to be trained anA made operational at the various departments.

103 5. ISERST's outreach activities

5.1. Royal Scientific Society The Royal Scientific Society institute of Jordan is the national research of that country. In mid-1992, contact was established with them in relation to a project to manufacture ump in Djibouti a mechanical wind that had been designed and tested by the RSS. hereafter, it is envisioned that technical transfer projects be organized in relation to other aspects will energy conservation. of renewable energy and

5.2 PANGIS. The PANGIS ( Pan African Network for a Geological Information S stem/ Roseau Pan-Africain pour un SystLme d'Information G ologiques) is a group of European and African documentary centers working in the field of African geology and linked for the exchange of information. to each other following a request made by admitted ISERST, the documentation center was to become a member of the PANGIS network in early, 1992. With the help of a PANGIS documentalist, a computer, printer, together with CDS/ISIS software have been installed documentation center to create in the operational. the database storage, which is now one of the ISERST documentation center personnel month of preliminary has received a training course at CIFEG ( Centre International de formation et d'Echanges an GLosciences/ International Center for Training and Exchange in the Geosciences), and is now head of the documentation center.

104 6.2 Recommendations for Future VITASAT Development 6.1. The information needs of the science and technology centers in Africa must be identified and provision made to provide this information on an ongoing, low cost basis. 6.2. Adequate investigation of the "computer infrastructure" of an area must be made before deciding whether or not VITASAT is an "appropriate" technology. As we gave seen in northern Somalia, where VITASAT repairs and/or troubleshooting are extremely difficult to arrange and the climate is harsh, it will be difficult to maintain an operational system for a long time unless the user is highly trained. 6.3. Adequate levels of training_ programs should be sought (es ecially in the computerbased network) for the technical staff to enable them to operate and maintain an operational system for long time use.

105 '!'U T r:l.l .E DE LA F )t !"PUBLl QUE

COtiSF.:7L

]) 1:7N7] NJ S''T I tATI CAN

iI ii ii ii 1 ii CO):P'PA.Bli r ------DIRECTEU R 1111;FC'10111 ------AUJOINT Q1IN' ------CH1.11Y DE SPRUCE A7)}:1;iI:;Tltk'Pl F

--.------_------!

I r------.....------) ------..-__..-..-_..r------1 I I 1 1 1 1 1 1 1 I 1 i 1 1 1 I

EILSICTIO1r SClYNC) S SECTION SCIENCE-1 SECTION SECTION DF.S TECHN01, 07J: M, LA T)r)i1tE HU LA I N NS VIE

Fait d DJTROUT7t Le Vu at Approuv6 Le Direoteur de WISERST Le Direoteur de Cabinet du Prdeident de la Upublique AVIS ARDALLATI

106 = Paper for the lateraatioaal Workshop at Digital Radio Technology aid Applications, is cooperation vith the lateraatioaal Development Research Ceatre; 24 - 25 AUG 92, Rairobi, teals. The opitioas expressed herein are eatirely the aithor'i aid sot necessarily those of CARE lateraatioaal or 6atelLife.

PACKET RADIO IN MOZAMBIQUE: ONE FAILURE ONE SUCCESS

Phil Gray, KA7TWQ/C9RPG CARE Mozambique 660 First Ave. New York, NY 10016

Abstract

Details of one HF Packet Radio network that failed after a four year attempt at implementation contrasted with one that was successful.

CONCLUSIONS

WHY HF PACKET MAY HAVE FAILED WHY HEALTHNET MAY HAVE SUCCEEDED

The concept was unknown in Mozam- The Ministry of Health was aware bique. of both the concept and of Satel Life.

No Mozambican of a high enough We had three men at Assistant level was involved to assist with Ministerial level to help with Telecommunications and customs TDM and customs difficulties. problems.

My limited experience with HF and An expert was on site to do the how sensitive it was compared to satellite station installation. the VHF I was used to working.

My lack of understanding how one The equipment was a donation and obtains money/purchases from CARE free (providing we could import headquarters. it without charge).

Not knowing the custom details Having a man on CARE staff who of importing electronics from understood local import regula- Canada to the United States. tions and procedures.

At least two separate and distant Only one station required. stations were required for testing, but there was only one qualified person available to do so.

107 CONCLUSIONS (CONTINUED)

HF HEALTHNET

The antenna configuration of di- No problems arose with the omni- poles was not optimum for HF. directionals. There was no money incentive. There would be a huge savings in international communications.

Telephone land lines improved the The BBS would augment and expand last 18 months of the project, so the station's reach up-country. the BBS emerged as a substitute.

No locals confident enough with Several computer literate and either radios or computers to interested doctors from the med- want to learn the operation. ical school.

Not enough free time to devote to Adequate free time plus a very the tests and training required. personal interest.

HF PACKET RADIO PROJECT: MARCH 1987 -- APRIL 1991

I joined the CARE mission in Mozambique, Southern Africa in January, 1987. A nation-wide food distribution project was in process in that big, long country -- much of which was inaccessible due to the civil war in progress. The country was a logistics education for even the most seasoned. We were the logistics branch of the governmental Department for the Prevention of Natural Calamities (DPCCN). The telephone system had rapidly deteriorated since the 1975 revolution displaced the Portuguese. Communication up-country in early 1987 was accomplished 90% by HF radio, and that was limited almost entirely to the ten provincial capitals.

All this brought to mind the experiment in Ethiopia the year before in which CARE and the Volunteers In Technical Assistance (VITA) had conducted fairly successful tests with HF Packet.l I asked for and studied those results. Three weeks into my work, it seemed there was a strong case for HF Packet. I wrote a concept paper to the CARE Country Director and thus began an exciting but overstating four year campaign.

The key people over the period were my two Country Directors, the East Africa Director in CARE New York, Dr. Gary Garriott of VITA in Washington DC, a member of the CARE New York MIS section (an amateur radio operator and acquaintance of Dr. Garriott), one consultant from South Africa and myself. Always in the background, of course, and vital to any national communications operation anywhere was the government; in this case, the Ministry for Telecommunications (TDM).

108 My first Country Director felt the idea was far too sophisticated and technical for Mozambicans to grasp at the time. But enough national and expatriate "staff were so positive that I was encouraged to write up a preliminary plan and budget: $30,000 for the ten provinces of the country. (The inexpensive estimate was due to the fact we already had the necessary CODAN Model 7727C radios on site.) For their information, I sent a copy to CARE headquarters in New York.

Our new Country Director arrived in July and was much more open to the idea. Suddenly, in September, the CARE New York East Africa director wanted more details and another budget. Then the Ford Foundation sought information. But the most positive news came from the Dutch Embassy: they wanted to spend some money on communications at the District level: radios to connect with the Provincial capitals.

The 17th of September, 1987, I submitted a plan for a district radio network to connect 50 of Mozambique's 70 districts. In the proposal was the $30,000 for a Packet Radio net with one station to be set up for satellite work. Copies went as always to New York. Two weeks later my MIS ham friend in New York reminded me that HF Packet was a very difficult medium in which to work. I forged ahead. I began to explore how to get the amateur frequencies from TDM for the satellite station and how to obtain the amateur license to work them. There were no known amateurs in country and the last license had been granted by the Portuguese government prior to 1974.

I left for home in February. I visited Gary Garriott at VITA HQ while I was in Washington in February. We discussed the project and possible VITA assistance. My equipment in Maputo was not adequate to do an acceptable job of integrating packet stations for what HF requires, but VITA could do it. I 'returned to Maputo in late March and the Dutch letter of approval was on my desk: $259,709 -- $30,000 for packet! We were on our way.

It was a busy Spring: meetings to define and assign tasks; a search begun for a Mozambican radio counterpart to work next to me; more frequencies requested from TDM; letters to AMSAT for advice in setting up the satellite station; a letter to the University of Surry, England volunteering our site in one of their experiments; VITA asked for a response to their offer to help buy and integrate the packet hardware; intensified requests for an amateur license.

On the 30th of May, we told VITA to proceed with the purchase and integration of ten packet stations. While that was going on, the National Director of the DPCCN was asked by my Country Director to assist with obtaining my amateur status. We held more meetings with TDM for additional frequencies for the district radios; little progress. Finally they were granted ten weeks later in August. About that time VITA also informed us they were still waiting for their money from CARE New York so they could begin -- over two

109 months waiting! In the meantime, the computers were ordered. Lastly during this period, a call came from the World Bank in Washington regarding the possible Beira/Maputo connection. Meanwhile, more telexes were passed to and from VITA and New to and from York regarding.payment for the hardware. It was both embarrassing and frustrating. But I had to learn the CARE routines.

Hoping to get some ideas and inspiration from an on-going Lesotho Packet project, I flew therein mid-November. While I met several Telecommunications officials and a few consultants, their project was no farther along than ours. Less, actually, because we at least had some of the hardware in country -- the computers.

When I returned to Maputo, there was a memo from Gary reporting the 7727C was no longer in production. Instead, the up-grade was an 8727 with similar specifications. (CODAN assured VITA the only things changed were improved electronics. Of importance to us were the cable and connectors which they said had not been altered.) We had no choice but to go with it. When I approved the change I didn't realize how it was to effect the duration, and hence, the ultimate end of the project. We learned all CODAN radios for Africa come from Australia. Any that go to the United States enter from Canada. As a consequence, the radio took over three months to get to VITA. And then testing was delayed because the microphone plug was non-standard and not available locally. This resulted in a further delay to make the required cables.

My Mozambican counterpart, Zita, was hired a few days before the year's end. Tie took over the battle for more and better frequencies for the district radios from TDM as well as my amateur license.

In mid-January I was in Washington to see both Gary about the hardware and the World Bank about the Beira potential. While excited about the World Bank interest, we were both distressed about the problems getting the 8727 in from Canada; it had yet to arrive!

By mid-March the project was two years old and not one packet had flown. The MIS ham in New York had transferred to Manila where he monitored our progress. He wrote that CARE Sudan was very interested in their own system and he suggested all sites should standardize.

On Monday afternoon, 24 APR 89, Zita and I opened the crates from VITA: ten Terminal Node Controllers (TNCs), cables, connectors, a new model CODAN radio (8727), and a Zenith 184 LapTop. An accompanying letter from Gary explained the contains and guaranteed all parts were in working order when sent.

As all this was progressing, I discovered to my great delight an amateur radio operator in town with a very strong interest in Packet Radio: Jerry. He was new in town but had already began work on assembling his own Packet Radio station from a kit he purchased in America. There was another ham in town, in fact the only licensed

110 amateur in the country, Kjell (C9MKT), but he was not interested in Packet.

I took Zita to Jerry's house. I wanted him to see the physical make up and connections of a Packet System and also to meet Jerry who had agreed to help us with the tests. The three of us made plans for test broadcasts the next time Zita and I traveled up-country. We would broadcast and Jerry would listen. Being unlicensed in Mozambique, he could not broadcast back to us. I would use all four channels available to us, Jerry would make notes and otherwise tell us how it went. This would be invaluable information for starting CARE/DPCCN's nationwide network of Packet Radio Stations.

Our first HF testing opportunity was from Chimoio which is 750 Km or about 500 miles from Maputo. But the first night there we got locked out of the radio room. I thought about poor Jerry struggling in Maputo to find us on the air. There was nothing we could do to tell him.

The next night we prepared both the old radio and the 8727C that Gary sent. We checked the voltage of the battery running the old radio: at just little better than 1OV, it was not the 12V we wanted but hooked up everything to it just the same: old radio, new radio, and the TNC. But we found to our great dismay the TNC was dead. The fuse had blown. Knowing full well fuses are impossible to find in the capitol of Mozambique much less Chimoio, we set about fashioning one out of a strand of copper wire and the old fuse itself. It worked and the TNC came on. All the hardware was ready to try by 19:45. We had an hour and fifteen minutes before Jerry expected us on the air.

Then I started running through the various communication software which would get the TNC to talk to the computer. None that I tried would give me the familiar "cmd" sign on the screen. All the software was more sophisticated than I needed at that particular moment. I wanted to see the old familiar commands I was used to with original TAPR TNC's. They were still there, but I couldn't get at them. I was becoming frantic as 21:00 came. I switched from trying Packet on the new radio to voice. It was at that point we discovered a real and unfortunate difference between the new and old models: the mic connections were the opposite on the new. That meant we had to exchange every female plug for a male one on the cable between the TNC and the old radios. I could only hope the CODAN agent in Maputo had some.

At any rate, using the old radio, I broadcast into space that we were having computer problems. I hoped Jerry was listening. As per our plan, I switched frequencies every ten minutes and each time I announced by voice that we were still experiencing computer difficulties. By 22:10 I had figured it out. We picked up from there and finished the hour on packet out as planned. Unfortunately, Jerry was on the first night (we were not), didn't hear anything so didn't get on the next night (we were).

111 As the weeks went by, I continued a few inconclusive tests using the 8727 while getting the connections altered on the older model. I notified Gary of that on 16 May. By 04 Jul, we were still in trouble with the 7727. My frustration at this point was enormous because I knew this same equipment had worked in Ethiopia three years before. On the 12th, I faxed Gary a pin out schematic to see if he could see an answer I had missed.

Richard Whiting of the Ethiopian experiment assisted Gary in a reply. Zita, the CODAN technician and myself all worked hard on it but we could not resolve the problem. I needed help and Joburg was closer than Washington. I called Peter Strauss -- a well known amateur in South Africa and a computer/communications expert. CARE decided to hire him to get us going.

August went by as discussions and details of Peter's visa, test equipment, etc. were worked out. As luck would have it, Gary was coming to Lesotho and a sort of race between him and Peter developed as to who could get over here to help first. But Gary lost out to the PACSAT launch in South America. Peter arrived the end of September, 1989. He worked here for two days with no discernible result. He suspected frequency and antenna problems. He took one of the PK232 TNCs back with him to test. He couldn't get it to respond to any tests until I suggested he use an amateur call sign; then Lt worked perfectly. He returned the 23rd of October with his own station (and accompanying visa and customs headaches). Our highest frequency was adequate for the 800 KM distance to Joburg, but the link was not strong. We traveled to two up-country sites with poor luck. He recommended a beam antenna.

While there was a flurry of activity with Peter's presence, there was not too much result for the money. Both my supervisor and the Country Director were getting anxious. They wanted a clear Scope of Work and a report. Peter's report arrived on 02 NOV and basically had two recommendations: 1 -- Change base antennae to beams, and 2 -- Get higher frequencies than our present maximum of 10 MHz. Of the two, at least we had some control over antennas. Getting different frequencies from TDM had been nearly impossible and we couldn't hope to do it again.

1990 did not start out well when the VITA computer was stolen off my desk while I was up-country. The software was backed up, of course, but we relied on both the machine and its MODEM. I got another machine going. Using some altered connections that Peter had sent us for the Motorola HT's, I wanted to try some VHF Packet. Our repeater failed before the tests. Also in February, Peter and I had some tantalizing attempts connecting with a Beacon from his station. So he decided to try once more -- for free -- to get us on the air. I saw him in Joburg on my way out for home leave in March and we made plans for when I returned. He was enjoying his flying lessons.

112 I saw the MIS man in Manila and saw Gary in Washington DC. Went with Gary to talk about Packet at one of the local amateur clubs and spent some more time at the World Bank -- more form and still no substance after three years!.

In early May, we resumed HF tests between Mozambique and South Africa. We were excited and hopeful about some successful connects that were made.. But in the next few weeks four things happen that took the heart out of the packet project and it never really recovered. First, and most tragic, Peter crashed and died during one of his flying lessons. Second, Jerry left Maputo for a new and we post. Third, I got our International EMAIL system working became connected into the rest of the CARE missions in the world. and Lastly, I installed the Remote Bulletin Board System (BBS) started the process of connecting with the ten provinces by land line.

The months ticked by, but I didn't realize by how much until I received Gary's telex in October asking me about the progress of the project. This probably come about because of David Balson's scheduled visit to Mozambique late November. I sent Gary a rather apologetic reply. Disappointed, but ever the positive fellow, Gary informed me VITA was now looking for,ground stations around the world -- would we be interested?

While I was home in January, 1991, two VITA representatives stopped by our office in Maputo (among several others). They met the Country Director. The Packet project.was not dead as far as VITA was concerned -- especially if we could also become part of the up-coming PACSAT experiment. Apparently, the VITA men stirred up quite an interest in town. By late February, Packet, at least in regards to satellite work, was on everyone's mind. On March 4th, Gary sent me a rough budget for a VITA consultant. Some encouragement from the MIS ham in Manila got us to early April. International Development Research Centre was on the periphery with interest and money for SatelLife, so I thought the two projects might be combined and save money.

After one or two more meetings with the Country Director and the other managers, I was able to send Gary the great news on 06 APR 91, that we wanted a VITA consultant to come and help me put in a station in the capital and one up-country. I sent a copy of this to Manila for their information.

I thought we had about $9,000 budgeted. On 24 APR, Gary had a man for the job and sent us a costing of $16,000. Unfortunately, on that very day, the MIS man in Manila sent us an EMAIL suggesting HF Packet was just too sensitive and slow for binary transfers, so we might want to reconsider our plans. The Country Director dropped it flat and didn't want to discuss it further. Four years of promise and frustration had finally gotten to him. And we didn't have but $4,000 available after all -- not enough to cover the air fare! My 27 APR EMAIL to Gary was full of disappointment and apology. Gary's

113 shocked response was one of disappointment equal to mine, but it was the end of the Packet project for Mozambique.

Note: After all the problems I had with the 7727, in early 1992, CODAN sensed there was a market and began to sell their 8525-B SSB transceiver fitted out with a CODAN MODEM and a LapTop computer for a "Data over HF System."

SATELLIFE (HEALTHNET) PROJECT: DECEMBER 1987 -- JULY 1992

I first came to know of the SatelLife Project when I read about it in the 02 Nov 87 issue of Amateur Satellite Report (ASR)z. According to the article, the originator of the concept, Dr. Bernard Lowe, was at the moment attending a conference in Moscow called the "Space Future Forum." There had been a definite and close cooperation between Russian and American doctors since its inception in 1985. By December, 1987, I was on their mailing list as a volunteer willing to participate in their activities however I could.

They were mentioned again in the 18 APR 88 issue of ASR3. This time HEALTHSAT-1 was their objective and they had hopes of a mid-1989 launch date. An update from the Coordinator, Dr. Malcolm Maclure, arrived at my Maputo office in May, 1988.

From May, 1988, until June 1990, I heard nothing from Boston. Then in the June, 1990 issue of 73 Amateur Radi04 appeared a report describing SatelLife and naming Dr. Charles Clements as the new Executive Director. I wrote him on 15 July 1990, to reintroduce myself and my location. He wrote back in early December to say things were moving very fast and several agencies have gotten involved with them including VITA and IDRC. Mozambique was ideally suited to be part of the project if we wished. In fact, SatelLife's field director in Africa, Mr. Mackie McCleod, had even met with the Minister of Health in Maputo.

VITA had not been idle during this period. Through their efforts, the Agency For International Development (AID) in Washington DC issued a telex to all their missions world-wide regarding the upcoming demonstrations on PACSAT and soliciting suggestions on its use. This telex reached my desk a week after the 20 JUN 90 deadline for input, but I immediately contacted other Non-Governmental Organizations (NGO) in the city to see if there was either knowledge or interest in this idea. Result: not much.

I replied to Dr. Clements' December, 1990 message in late February, 1991, saying I would pursue the required permission for frequencies and having the Minister of Health on our side would be most beneficial.

114 At this time also I began contacts with the AID branch in Maputo, dredging up their year old telex about PACSAT in an effort to get help acquiring frequencies and perhaps even a little money.. I spoke as well with the Canadian official representative here regarding David Balson's 21 MAR 91 fax informing me of the IDRC money for was satellite experimentation. I met also with Dr. Julie Cliff who overseeing several projects for the Ministry of Health. to meet in In June I was in the States. Dr. Clements and I tried Washington after my discussions with VITA, but somehow we missed each other. Two months passed after my return to Mozambique and 09 SEP 91 issue of then I read a small mention of SatelLife in the OSCAR Satellite Reports which accompanied a photo of Vern Riportella but holding an antenna tested by SatelLife. I tried to fax them their numbers had changed.

So, unknown to me, on 07 NOV 91, Dr. Clements sent an introductory letter to Dr. Jorge Barreto of the National Institute of Health (NIH) in Maputo. He spoke about SatelLife, IDRC, engineer Edson Pereira (who was installing ground stations in the region) and the offer to donate a satellite station to the NIH. and she I saw Dr. Cliff on the street not too long after that informed.me of the message from Dr. Clements. I went immediately to do meet Dr. Barreto to discover what had transpired and what I could and to help. Though we had never met before,.we got on very well devised a strategy to obtain the required licensing and frequencies. We also needed to get the donated equipment into the country as cheaply as we could -- free if possible! Using their new FAX number, on 21 NOV, I informed SatelLife what was transpiring.

By return FAX Dr. Clements informed us the engineer had to be ticketed for Maputo within four days if we wanted him to fly to Maputo to assist us with installation. And instructions for the hardware had to be sent to Boston immediately. On the shipping we 26th of November, we held another strategy meeting. This time involved the Chairman of the University Medical School and the Director of the General Hospital. It was arranged to have the equipment addressed to the hospital -- which would allow it to enter duty free. Dr. Barreto was somehow able to use his influence and connections downtown to obtain licensing on the condition the station was located at the medical school. I prepared to get a visa for the engineer but IDRC was doing that in Kenya.

While we now knew where to send them, we were not sure what documents we needed to insure duty free importation. For this we got help from my people at CARE. The Mozambican in charge of CARE's procurement told me what kind of documents we should have available to present to customs officials at the airport. I requested these from SatelLife on 09 DEC. They were immediately sent to us by FAX. On the 11th the cargo arrived at the airport and the usual customs hassles began. More on that later.

115 On the 21st Edson arrived -- I figured he was the one with coax cable wrapped around his shoulder! It was Saturday and we could not get into the medical school until Monday.. I invited Edson to dinner and asked Don Findley if he would like to join us. Don and I had been planning some VHF packet experiments for the past few weeks but hadn't finished setting up the equipment. He was most interested in being a part of this satellite venture. So Monday morning we all met at 08:00 ready to go to work and only six days to do our magic.

The three of us met first with Dr. Barreto to introduce Edson and explain what we would be doing for the next few days. Then all of us went to the Medical School to meet with Chairman of the Faculty and the Director of the General Hospital. We all noted the equipment had not cleared the airport yet so the pirector send his men to customs to find out why. Then we did a tour of the building to determine the best location for the station. Dr. Barreto and Edson needed to meet for a while so Don and I took this opportunity to go visit TDM and check on frequencies and.my amateur license. For four years I had not felt it so necessary to have my license that I needed to do anything out of the ordinary to obtain it -- a few informal lunches, the occasional stop by the offices for coffee, etc. But we had to have that document now. I carried with me a bottle of expensive whisky. In ten. minutes I walked out with my new amateur license: C9RPG.

After lunch we began to worry about customs. Still no equipment was cleared. Customs bureaucrats everywhere.in the world simply hate duty free imports, and the personnel here were no different. They had stalled for nearly two weeks. It was time to bring our the heavy artillery: By day's end Dr. Barreto was chosen to go to customs the next morning. He left at 08:30. Hoping to help by just being there, Don, Edson and I got there at 10:00. We walked in and just stood behind Dr. Barreto who was seated arguing with some official. It seemed to work. A few minutes later Edson left with the official for the airport and an hour later he drove up to the Med-school with all the boxes.

Edson's time with us had been reduced from six days to little over three. We started work immediately and continued through Christmas day. At 15:30 on December 26th, our Christmas present arrived as we sent our first message: #91122601 addressed to Jon Metzger at SatelLife/USA informing him that we were up and working on UOSAT 141 Edson left the next"day for South Africa where the head of South Africa AMSAT, Mr. Hans Van de Groenendaal received him as an overnight guest.

In the next few days Don and I met and trained professor Humberto Faustino who was to be our counter-part and interface with the medical school. Dr. Faustino had studied in Australia and was very excited about the prospects of augmenting the med-school library through the SatelLife project. We were reminded by both SatelLife and VITA, however, that we were on amateur frequencies and must therefore limit transmissions to only what is permitted on those bands. So we worked on setting up basic systems of 116 daily/weekly/monthly chores for operators, message tracking, etc. and tried to show off the station to all who were interested and many who weren't.

Then in February of this year, all amateurs and SatelLife stations were traumatized when the great switch between UOSAT3 and UOSAT5 took place. But since then, we have mostly been waiting to use. the SatelLife station as it was designed: to pass medical information and research back and forth among the users. Also in the works for Mozambique is a land line BBS that will connect the Maputo station to health centers in the provincial capitals. CARE operates one such BBS now for six sites which phone lines are of the quality required. The phone situation improves at a steady rate at the remaining four capitals and we expect them to be,operational by year's end. In line with that, an IDRC representative from Kenya, Frioze Manji visited Maputo and instructed us to get a users group going so we can better use the station and the up-coming BBS. Our first meeting is 12 AUG 92.

References

1R. A. Whiting & H. P. Snyder, Report and Recommendation Packet Radio Feasibility Demonstration Project 6 June 1986. (Washington DC: VITA, 1986). 2Moscow Space Forum Highlights Future Space Activities, Amateur Satellite Report (November 2, 1987; No. 161): page 1.

3Conferees Agree In Principle To Deploy PACSAT From Mir, Amateur Satellite Report (April 18, 1988; No. 173): page 1.

4SatelLife Packet, 73 Amateur Radio, (June 1990): page 7. 5PACSAT Proliferation, OSCAR Satellite Report, (September 9, 1991; No. 228): page 2.

117 118 Satellite-based Developments

119 120 UNWERSffY OF SURREY I SST

TECHNOLOGICAL UNIVERSITY Compact Campus Located 35km South West of London 20km to London-Heathrow and London-Gatwick Airports 4750 Students in Science, Engineering, Linguistics

MULTI -DISCIPLINARY SPACE RESEARCH i~N Spacecraft Engineering (Electronic Engineering) Ir Satellite Communications (Electronic Engineering) Space Structures (Civil Engineering) Space Materials (Materials Science) Space Robotics (Mechanical Engineering) SURREY CENTRE FOR SATELLITE ENGINEERING

TECHNOLOGY TRANSFER INTO INDUSTRY Surrey Satellite Technology Ltd UNWERSOTY OF SURREY / SSTL SURREY SATELLITE TECHNOLOGY Ltd

SSTL is a wholly-owned company of the University of Surrey working intimately with the UoSAT Spacecraft Engineering Research Unit. The Company's objectives are to: Transfer small satellite technology from academic research labs to the industrial sector for commercial exploitation

Undertake industrial research & development contracts - small satellite platforms - payloads - groundstations & orbital operations - training & education Provide funding for the: Centre for Satellite Engineering Research (100 personnel) at Surrey, which aims to become a European academic centre of excellence in satellite engineering. UNWERMY OF SURREY / SSTL LOW COST MICROSATELLITE MISSIONS.

Trend of space activities has been towards: - larger, more complex, multiple missions - increased costs - longer timescales - 'eggs placed on one basket' - major (expensive) groundstation facilities - international projects - increased bureaucracy

'Small' satellites can provide a complimentary role to 'large' missions for certain well-focussed objectives: - within short timescales (12- 18 months) - low cost (£ 1.5 - 2 million) - good reliability (8 years operation) - sophisticated functions - small teams, modest facilities, direct access UNOVERS ` Y OF SURREY ® SSTL

CLASSES OF SATELLITES Class Mass

Large satellites >1000 kg

Small satellite 500-1000 kg Mini-satellites 100-500 kg 10-100 Micro -satellites T kg

Nano - satellites < 10 kg UNIVERS ff Y OF SURREY / SSA u i!

APPLICATIONS OF__MICROSATELLITES

Technology Demonstration

Specialised Communications

Small - Scale Space Science

Remote Sensing

Education & Training yr1 1_1 ( J ® C-ZS U L UoSAT MICROSATELLITE PROGRAMME

THE UNIVERSITY OF SURREY HAS PIONEERED MICROSATELLITE TECHNOLOGIES THROUGH RESEARCH INTO COST-EFFECTIVE SATELLITE i. ENGINEERING TECHNIQUES.

THIS HAS FOCUSSED UPON THE DESIGN, CONSTRUCTION, TEST, LAUNCH AND ORBITAL OPERATION OF MICORSATELLITES IN LEO.

SEVEN MICROSATELLITES HAVE BEEN BUILT BY SURREY: Launched Status UoSAT-1 1981 (NASA) Re-entered 1989 UoSAT-2 1984 (NASA) Operational UoSAT-3 1990 (Ariane) Operational UoSAT-4 1990 (Ariane) Failed (after 30hrs) UoSAT-5 1991 (Ariane) Operational KITSAT-1 1992 (Ariane) Being commissioned S80/T 1992 (Ariane) Being commissioned FOUR NEW MICROSATELLITES ARE UNDER CONSTRUCTION: PoSAT-A 1993 (Ariane) Portugal HEALTHSAT-B 1993 (Ariane) SatelLife KITSAT-B 1993 (Ariane) KAIST (Korea) CERISE 1994 (Ariane) Alcatel (France) U®SAT / SSTL STORE & FORWARD COMMUNICATIONS MICROSATELLITES

THE UNIVERSITY OF SURREY HAS ALSO PIONEERED DIGITAL STORE-&-FORWARD COMMUNICATIONS TECHNIQUES WITH VITA & SATELLIFE USING INEXPENSIVE MICROSATELLITES IN LOW EARTH ORBIT TO PROVIDE EFFECTIVE COMMUNICATIONS TO REMOTE OR DISASTER- STRICKEN AREAS WORLD-WIDE:

UoSAT-2 (Digital Communications Experiment) The first microsatellite to demonstrate modern digital store-& -forward communications in the Amateur Satellite Service & with VITA stations.

UoSAT-3 (PACSAT Communications Experiment) Experimental microsatellite using advanced packet communications techniques in the ASS and for pilot demonstrations by VITA on non-ASS frequencies. UoSAT-5 (PACSAT Communications Transponder) Operational S-& -F transponder for use by SatelLife on non-ASS frequencies.

UoSAT-5 & 5 OPERATIONS WERE SWITCHED FOR OPERATIONAL CONVENIENCE IN MARCH 1992

UoSAT NCROSATE U E SYSTEMS

SOLAR CELL ARRAYS ANTENNAS

UPLINK DOWNLINK RECEIVERS TRANSMITTERS

DEMODULATORS MODULATORS

BATTERY MULTI-PX MULTI-PX CHARGE SWITCH SWITCH u REGULATOR

TELECOMMAND ON-BOARD TELEMETRY DECODER COMPUTER ENCODER NiCd POWER CONDITIONING BATTERY MODULE RANDOM ACCESS TO SUBSYSTEMS MEMORY FROM SUBSYSTEMS

POWER DISTRIBUTION MODULE ATTITUDE SENSORS PAYLOADS SUN EARTH mrrrrr MAGNETIC -0 S'FaSv'S 0.10 S O( CONTROL BOOM EXPERIMENTS MAGNETS 1`AHOROSATELUTES FOR` COMMUMCAMONSeaee

Store-&-Forward Communications Transponder Satell-ife (USA)

To provide communications to medical & disaster relief teams working in remote regions world-wide.

\4 DOWNLDM

RI tie 80C186 processor + 13 MBytes SRAM 9.6 kbps AX.25 VHF/UHF communications small 'briefcase' ground terminals can store 3000 A4 page or 60 images Uv 8 AT PR0GiR A hAN E

UoSAT-5

LAUNCHED IN JULY 1991 BY ARIANE

- 775 km, polar, sun-synchronous orbit (ERS - 1) - mass of 50 kg

PAYLOADS HEALTHNET Transponder Payload (Satel ife) Solar Cell Technology Experiment (©RA) Earth Imaging System (SERC/SSTL) Advanced Attitude Control (SSTL/MMSF)

COMMISSIONED IN ORBIT WITHIN ONE WEEK

0 OPERATIONAL UMVERVTY OF SURREY / SST UoSAT-5 MICROSATELLITE MISSION

UoSAT-5 TECHNICAL SUMMARY

Mass . `` ka Dimensions 5 x 3,-5-4f) 6 78mm Solar Arrays 672 GaAs solar cells mounted on body Power 49 watts raw peak power 18 watts processed average over one orbit Attitude computer-driven electromagnets compiimenting Control 6m gravity -gradient stabilisation boom Sensors 2x magnetometers, Sun sensors, Earth Sensors Telemetry 64 analogue channels 144 digital status indicators Telecommand 4 redundant micro controllers 120 latched command signals Computers 8-MHz 8OC186 primary on-board computer 13.5 MBytes PAM, 72 bits parallel l/0 4 channels multiprotocol I/0, Ouadron 4-MHz Z80 secondary OBC, 150 kBytes RAM 20-MHz T800 Transputer parallel processors 2.5 MBytes PAM (16k SOS) U®SQT / SSTL STORE & FORWARD COMMUNICATIONS MICROSATELLITES

HEALTHSAT-B IS BEING CONSTRUCTED BY SSTL AT THE UNIVERSITY IN II OF SURREY FOR SatelLife AND IS SCHEDULED FOR LAUNCH SEPTEMBER 1993.

HEALTHSAT-B WILL PROVIDE DEDICATED S-&-F COMMUNICATIONS FOR STATIONS USING SMALL, PORTABLE GROUND TERMINALS IN THE HEALTHNET SERVICE, OPERATING AT VHF/UHF.

HEALTHSAT-A Basic Specification: Two 80C186-based microcomputers with 0.5 M-Bytes RAM Two 16 M-Bytes RAM-Disks Redundant downlink transmitters with output power adjustable up to 7W (+ 8.4dBW) 9.6 kbps / 38.4 kbps downlink data rates

0 Three uplink receivers at 9.6 kbps data rate "ACROSAATELUTES FOR REMOTE SENSN Gave

HIGH-DENSITY, AREA CCD OPTICAL DETECTORS COUPLED TO POWERFUL ON-BOARD PROCESSING ENABLE MICROSATELLITES TO PROVIDE MEDIUM-RESOLUTION REMOTE SENSING. UoSAT-5 CARRIES AN EARTH IMAGING SYSTEM (EIS) COMPRISING: 578 x 576 PIXEL CCD SENSOR GIVING GROUND RESOLUTION OF 1.5-2 km IMAGE DIGITISED TO 256 LEVELS ORANGE FILTER FOR ARID/VEGETATION & LAND/SEA

DATA PROCESSED BY 2xT800 TRANSPUTERS AT 20MHz

DATA TRANSFERRED TO 13 MBy to RA MDI SK VIA LAN IMAGING PROGRAMME SCHEDULED AUTOMATICALLY VIA 80C 186 OBC MULTI - TASKING REAL-TIME OPERATING SYSTEM 60 IMAGES CAN BE STORED IN THE ON-BOARD RAMDISk FOR DOWNLOADING TO GROUNDSTATIONS USING 10kbps AX.25 PACKET COMMUNICATIONS PROTOCOLS AT UHF VITA Operations using UOSAT-3

Eric Rosenberg, WD3Q Satellite Communications Specialist VITA

and

Gary Garriott, WA9FMQ Director, Informatics VITA

1. Introduction to VITA

Volunteers in Technical Assistance is a not-for-profit organization dedicated to providing technical and project assistance to developing countries. For more than thirty years VITA has provided by-mail as well as on-site assistance through its Inquiry Service and long-term overseas programs. A small central staff headquartered in Rosslyn, Virginia, is supplemented by field personnel in eight countries as well as a volunteer roster of over 5000 highly skilled technicians and engineers. VITA has published over 200 technical papers and books dealing with a wide range of development applications from agriculture to water supply, generally emphasizing low- or medium- cost technologies that can be locally produced and maintained in rural settings.

II. History of VITASAT Program

In 1980 VITA began to experiment with alternatives to the by-mail diffusion of technical information. At that time, about 25 two- way real-time audio teleconferences were held over the PEACESAT network. PEACESAT was made up of user groups throughout Oceania and the South Pacific using relatively low-cost terminals on the ATS-3 NASA satellite. While this experience revealed the power of interaction between users and VITA volunteers, it also pointed out the difficulties in effective real-time links over many time zones and the possible advantages of asynchronous communication. VITA also began doing semi-regular broadcasts over the Voice of America on technology overviews at this time.

In 1983 VITA contracted with the Radio Amateur Satellite Corpora- tion (AMSAT) to develop a technical specification and design definition for a LEO PACSAT mission. This activity culminated in a meeting held at Wang Laboratories late in 1983 at which the University of Surrey offered to make its UOSAT-2 bus available for a "Digital Communications Experiment" package if it could be

135 readied in time. Over the next six months AMSAT and VITA volunteers, coordinated by a VITA-paid consultant, managed to complete the DCE and integrated it into UOSAT-2. It was subsequently launched by NASA in March of 1984 and continues to function to this day. VITA's amateur DCE station was established in May 1986 and has been used for occasional demonstrations and training sessions.

The DCE was, however, inappropriate for regular communication regarding VITA's development "business." Subsequently, an effort was undertaken to acquire an experimental license on "non-amateur" frequencies to be able to take advantage of upgraded technology as well as to introduce store-and-forward communications to a limited number of demonstration stations in developing countries. Funding was also acquired from a variety of US government and non-government sources to support the design of the "PACSAT Communications Experiment" providing in-orbit testing as well as two "production" PCE-type payloads for a future operational, dedicated satellite. VITA also purchased satellite transmitters and receivers using the experimental frequencies.

The PCE on UOSAT-3 was subsequently developed by Surrey Satellite Technology, Ltd., and successfully launched by Arianespace in January 1990. It was commissioned early in 1991 and has provided over 100 amateurs from almost 30 countries with many exciting moments and fairly reliable communications to boot! Recently, operations on the experimental frequencies were initiated.

VITA's goal is to use the experience gained from PCE operations to develop the corporate track record and infrastructure to support a fully-dedicated satellite for launch sometime in 1993 (VITASAT-A) which could support at least 500 ground stations worldwide. A second satellite (VITASAT-B) could increase this figure to as much as 1000 with a spare on the ground. To this end, VITA has initiated a difficult (and expensive) process to get FCC authorizations and rules created to support a non-profit LEO store-and-forward satellite service, while simultaneously. working to promote adoption of similar allocations at the World Administrative Radio Conference to be held in Spain in early 1992. Table I lists those frequencies proposed to be used by VITA on both the PCE and VITASAT-A/B. VITA was invited to participate in the launch of UOSAT-F, but with the exception of a back-up capability negotiated with the primary US customer (SatelLife), opted instead to pursue the operational phase with its available resources (VITASAT A/B).

III. Objectives of the PCE phase, VITASAT Program

As noted, VITA hopes to gain useful experience from the PCE and its utility in real-life development and relief situations in Third World countries. To this end, it is negotiating Memoranda of Understanding with sponsors who are other non-profit organizations, educational institutions, government bodies and develop-

136 ment agencies responsible for acquisition of licenses/authorizations as well as operations. VITA plans to use its staff and volunteers--all radio amateurs--to perform on-site installations while simultaneously conducting operator training. Software is being independently developed and constantly upgraded and evaluated. Initial software development was. ithrough a contract with the Virginia Polytechnic Institute (Blacksburg, Virginia).

The first tier of installations will be made using the "fixed station" equipment complement as noted elsewhere in this paper. A "portable station" is also under development which will be deployed in selected disaster response and mitigation sites, such as refugee camps and AIDS epidemiology work in Africa.

VITA expects that 30-40 PCE demonstration stations will be established throughout Asia, Africa and Latin America with perhaps 50% of these in Africa. General topical areas are education, energy/environment and health.

Performance of groundstations in this demonstration phase will be evaluated by VITA using both technical and sociological criteria.

IV. Overview of VITA Field Sites

Besides providing for other organizations to use the PCE, VITA will also be using the system for at least two of its own projects. one will be based in Peshawar, Pakistan, where VITA provides cross-border assistance to increase agricultural production in Afghanistan. The present objective is to restore the Afghan agricultural infrastructure, which has been badly damaged by more than a decade of war and enforced neglect. Primary attention in this project is given to the repair of irrigation systems essential to the production of wheat and other crops needed to feed the people and of the roads and bridges making it possible to bring farm produce to market and deliver improved seed, fertiliz- er, and other farm inputs to the villages.

Another project is located in Djibouti, a small country on the Horn of Africa, where VITA has been working since 1982 with the National Institute for Higher Scientific and Technical Research. The goal in this project is to help the government reduce the country's dependency on imported fossil fuels. VITA has introduced conservation techniques and renewable energy technologies and helped develop a long-term national energy strategy. VITA also helped set up a permanent National Energy Council, revise building codes, and conduct a publicity campaign promoting energy conservation.

VITA is not the only non-profit organization planning a series of installations for its projects. An arrangement supporting other multiple installations has been made on a case-by-case basis with

137 other organizations. One of these is SatelLife. While a newcomer to the international development arena, SatelLife has been notably successful to date in its goals to promote LEO satellites for the exchange of health information. SatelLife expects to install PCE groundstations in a number of East African countries linking medical schools in the "ESANET" countries which at this date include Kenya, Tanzania, Uganda, Zambia, Zimbabwe and Mozambique. SatelLife's Memorandum of Understanding with VITA includes a "non-interference" provision between the PCE and UOSAT-5 on which SatelLife intends to be a major user. VITA and SatelLife are attempting to coordinate many of their activities since the similarity of their goals and use of the same satellite bus can provide system redundancy if and when required.

V. PLAN International Installation - Sierra Leone

PLAN had originally proposed setting up a ground station at their regional headquarters in Dakar, Senegal, and in Conakry, Guinea. In both instances, the local PTT's refused to grant permission. While Sierra Leone had been the third choice for PLAN, it was a fortuitous one for VITA, as the communications infrastructure in this, the poorest of African countries, is virtually non-existent.

For one who has not travelled to Sierra Leone, it is difficult to understand the degree to which PLAN's Freetown office is cut-off not only from its International Headquarters in Rhode Island, but too, the two offices it operates up-country in Makeni and Moyambo.

After years of providing intermittent commercial mains power in Freetown, none has been produced since mid-February. Those that want or need electricity and all it provides (pumping of water and sewage, refrigeration and air-conditioning) must generate it themselves. To that end, PLAN has two diesel generators, on 25 kva and a backup of 7.5 kva. The generators are only run during business hours.

The telephone system is at best erratic. As an example, the 280 outgoing telephone circuits at the US Embassy go only as far as the walls of the Embassy itself. Messages between the three PLAN offices are relayed by courier or HF (voice] radio. To originate international telephone calls -- and thereby send faxes or telexes, PLAN is at the mercy of SLET, the Sierra Leone External Communica- tions Company. A call is placed to the overseas operator, who is given a list of the overseas numbers to be called. Later in the day, the operator calls PLAN's office with the connection(s).

As can be imagined, this method of calling overseas is not terribly reliable. Fax and telex service are especially poor, the former dependent upon high quality telephone circuits, while the later is dependent on another set of operators who re-key the message and send it out of the country.

To overcome.these difficulties, PLAN's Country Representative in Sierra Leone, Mohan J Thazhathu, took it upon himself to secure the

138 proper licensing necessary to establish a VITASAT station in Freetown. With UO-14 and later with VITASAT-A & B, Mohan envisions a time in the not-to-distant future when he can send his monthly reports, financial statements, and other adminstrative materials via satellite to PLAN's International Headquarters.

In addition to establishing the VITASAT ground station, PLAN also purchased three PK-232 terminal node controllers to further connect the offices in Sierra Leone.

As with any project overseas, a great deal of preparation went into the establishment of this station. Equipment from a variety of manufacturers had to be evaluated, and bid specifications prepared. Once a decision was made to go with a particular piece, it had to be integrated into the system. One particular problem we've run into is the difficulty in obtaining the Kenwood TS-790A transceivers. For unknown reasons, this radio has become particularly hard to obtain.

(A listing of the equipment we use for our installations is included in Table II.].

As requird, the equipment was modified by a vendor in Washington, DC. This included the installation of a VPI-designed buffer/summing amplifier between the varactor and FSK input, expanding the transmit and receive range of the radio, setting the deviation, and tapping the discriminator for the audio output.

Documentation for the installation of the hardware and use of the software had to be written, edited, proofed, (re-written) and collated.

The equipment was then inventoried and packed in two footlockers and two boxes for transport with Eric to Freetown. This was easier said than done. By flying on two different carriers -- one American, one Dutch -- we were subject to two different sets of rules for overweight baggage. After a false start which sent us back to re-distribute the weight of the footlockers, the overweight baggage tariffs were paid and Eric was off to Freetown.

Eighteen hours and three flights later, Eric arrived in Sierra Leone. The equipment arrived intact, and unscathed. His suitcase, on the other hand, did not appear for another eight days.

Installation of the antennas and related equipment was straight- forward. While the pace was slower than what he might have expected in the USA, no difficulties were encountered. Bringing everything from electrical tape to feedline and almost everything in between kept the project close to schedule.

The difficulties he ran into were with the items not checked beforehand. These were the items we never figured to have trouble with. We were wrong.

139 The boom sections of the 2 meter antenna would not easily connect together, so time was lost in Freetown as two fitted ends were filed down to mate properly (we've since checked the three antennas to be installed in subsequent projects, and all suffer from the same problem. Fortunately, the manufacturer was willing to replace the defective pieces].

When installing the Kansas City Tracker and tuner software (version 2.4.3], a bug was found in the INSTALL.EXE program. A frantic call to the US manufacturer on July 4th was made to remedy that situation.

And finally, Eric did not bring rotor cable with him. A large part of a day was lost in search of .a suitable replacement.

While the antennas and roof tower were going together outside, inside the office PLAN personnel set about hooking up the radio and installing the software on a NEC Powermate computer. PLAN has made a sizeable investment in computers in the Freetown office, and its staff is quite knowledgeable in their use. Installing the software and training the staff to run the station turned out to be a relatively easy task, with PLAN's staff as eager and intelligent students.

In preparing for this and future installations, we had decided that the initial operation of the ground station would take place on the amateur bands. This would allow us to check out the the equipment and give the individual installing the equipment an opportunity to use his free time to give out a 'new one' on the satellites.

On his arrival in Freetown, and with the intervention of Dave Heil, 9L1US, Communications Officer at the US Embassy, Eric was given permission to operate on the amateur bands as 9L1/WD3Q.

The Officer in Charge of Frequency Management for the Sierra Leone National Telecommunications Corporation is Cassandra Davies. Mrs. Davies, in addition to passing judgement on the viability of this project, is 9L1YL -- and President of the Sierra Leone Amateur Radio Society. During his time in Sierra Leone, Eric went to the Radio Society's monthly meeting and gave a short talk about amateur satellites, VITASAT, and the PLAN project in Freetown. In addition, the extra feedline from the PLAN installation was donated to the Society for it's newly established club station.

To say that the station worked on the amateur bands would under- state the point. This location stands alone in the satellite's footprint for the entire pass. A large number of transactions can and did occur between the satellite and the ground station.

Based on our experience at the VITA HQ station in Washington, Eric brought a 2 meter amplifier for the uplink. This proved to be unnecessary. What he was not prepared for was an extreme amount of interference on the VITA (429 Mhz] downlink. In a country that has no monitoring service, and without another radio capable of

140 receiving in the UHF frequency range, Eric was unsuccessful in tacking down the source of the interference. It is hoped that a series of interference mitigating solutions under consultation will allow PLAN to access the VITA frequencies.

VI. Expectations of the PCE

It is our hope that a viable communications network will be established among development agencies and organizations. To that end, we hope to establish as many as 20 ground stations in sub- Saharan Africa, Asia, Oceania and Latin America by the end of this calendar year. The next stations to go on line will be at VITA projects in northern Pakistan and Djibouti. Soon thereafter, we expect other stations to be established in West and Africa Africa.

The PCE on UO-14 is a service shared by amateurs and non-amateurs. Increased use of the non-amateur piece of the will, naturally effect the amateur side. Based on the present activity on the PCE, one may assume that the UO-14 user community in northern Europe, and North America could see little differences in the operation of the satellite. However, as the requests for the non-amateur use of the PCE grow, this may change. Considering that some of the files being up- and downloaded could be as large as 100k bytes, the satellite may not always be available when an individual wants to access it.

At the present time, the satellite's transmitter switches over to the VITA downlink frequency for between 250 milliseconds to five (5) seconds out of every twenty-five (25). This may increase over eastern North America as we interact with the satellite from our station in Washington. As the satellite will accommodate two users at any one given time, regardless of where the user comes in, we will be subject to the FULL flag as much as anyone else.

VITA software is under continuous evolution and development. With the current version, the user can create a list of transactions (uploads, downloads, directories) to be attempted in one session. A future iteration of the software will automatically request downloads of files over a certain size (most likely 8 kbytes) as broadcasts and not when in the connected mode. Likewise, uploaded files over 4 kbytes will automatically be compressed prior to uploading. Again, as we are in the demonstration phase of this project, we expect to modify the software to suit the actual, real- world needs presented by our user base.

VII. Now Individuals Can Cooperate with VITA

No low-earth orbiting satellite can exist without the help of the amateur satellite community. As we discovered in Sierra Leone, the local amateur radio community can be of assistance in supporting this project. Those of us working on this project have been involved in the satellite community and active on other amateur

141 satellites for years.

But we cannot do it alone. VITA is an organization that runs on volunteer power. We are always looking for qualified to work individuals with us on our various projects, in this instance up ground setting stations in the developing world. If you have UoSAT-14 station put a on the air, we would like to speak with you!

VIII. The Future -- VITASAT Program While VITA is currently in Phase Two of its three-phase program, have continued we to look forward to Phase Three -- construction and launch of a fully dedicated and operational store-and-forward satellite system (VITASAT-A/B), perhaps beginning as early The approach as 1993. taken by VITA's Board of Directors is to participate as fully as possible in domestic and international creating proceedings the service structure and frequency allocations necessary to make VITASAT a reality. Once at least a temporary authorization construction from the FCC has been acquired, negotiations will begin in earnest for a suitable launch opportunity. authorization Such an may be forthcoming within the next few months. To date VITA has filed an "Application for Authority to a Non-Profit Construct International Low-Earth Orbit Satellite System" and companion "Petition a to Establish a Non-Profit International Low- Earth Orbit Satellite Service" with the FCC. In addition, VITA has participated as much as possible in proceedings leading toward a favorable acceptance of Low Earth Orbit World satellites at the Administrative Radio Conference (WARC-92) to be held during in Spain February-March 1992. Partly due to VITA's work in domestic joint working group and industry advisory committees ITU's CCIR as well as the (technical communications committee), Low Earth Orbit satellites are on the WARC-92 agenda. While VITA's application technically describes a "fixed satellite service", we have proposals made our compatible with the "mobile satellite service" language to be considered by the WARC-92.

As this paper is written, the FCC has adopted a Report and recommending Order to the U.S. State Department that most of the 137-138 MHz band be allocated to LEO MSS on a co-primary basis with other services, with remaining LEO MSS use of this band on a secondary basis to the meteorological-satellite service; and that the 148-149 MHz and 400.15-401 MHz bands be allocated to LEO MSS on a co- primary basis. This is entirely consistent with VITA's proposals. No existing amateur frequencies are involved.

In summary, we believe that the VITASAT program has used resources available through amateur radio in an appropriate manner throughout the program. Phase one was a "proof-of-concept" experiment using amateur frequencies and volunteers, all of whom held amateur licenses. Phase Two still uses amateur radio volunteers and frequencies for testing purposes only, with "operations" on "non- amateur" fregencies (see Table I). VITA paid for the design of the

142 PCE which has contributed to its success to date for the amateur satellite-service. Phase.Three will no doubt still use the human resources available through radio amateurs, but in terms of operation will be totally separate from the amateur satellite service.

VITA acknowledges a debt of gratitude to radio amateurs as individuals and to the amateur radio service which has and is providing an experimental environment for the development and evolution of the VITASAT system. VITA has actively promoted amateur radio among the non-hams it has worked with and has consistently tried to extend its appreciation of the contributions made by amateurs to the non-amateur public. Many hams have become VITA volunteers and we solicit others interested in joining VITA to become volunteers as well. We are proud of our working relationship thus far with the amateur community and look forward to continued and expanded contact with those supportive of our efforts.

VITA Volunteer Hugh Pett, VE3FLL, described his commitment seven years ago upon the successful launch and commissioning of the DCE on UOSAT-2 this way:

"If we are able to put more food on somebody's table five years from now, we'd be extremely proud of our efforts."

While that dream is a little "behind schedule," it is still very much alive.

143 Table I.

VITA PCE and Proposed VITASAT Frequencies

UPLINK DOWNLINK

PCE 148.560 429.985 * 148.260 428.010 *

VITASAT ** "Solution A" 400.175 137.705 400.225 137.735 400.275 "Solution B" 149.825 400.175 149.855 400.225 149.885

* These frequencies are in bands classified as "shared" government and non-government (amateur) in the United States. Before the experimental application was submitted to FCC, consultation with ARRL was conducted and a determination made that any existing amateur activities would not be unreasonably disrupted.

** Solution A is fully compatible with a continuous coverage domestic LEO system using FDMA techniques (proposed by OrbComm). Solution B is fully compatible with a continuous coverage domestic LEO system using spread spectrum techniques (proposed by Starsys).

Table II.

Equipment used for VITASAT (PCE) Stations: 80286/80386 based computer VITA VGS software package (written by Dave Carre) Kenwood TS-790A dual band transceiver PacComm Micropower-2 Terminal Node Controller NB96 internal 9600 baud FSK modem VPI/PacComm Summing/Buffer amplifier board (for TS-790A) MZ Enterprises 2M-CP14 uplink antenna 436-CP30 downlink antenna L.L. Grace Co. KC Tracker Package for Yaesu 5600A Controller KC Tuner Option N4HY Quiktrak Yaesu KR-5600A rotors Astron RS-20A/220 power supply Astron RS-35A/220 power supply RF Concepts 2-417 amplifier Mirage/KLM KP-2/70cm Mast-Mounted preamplifier for 430 Mhz Angle Liner 429GN in-line preamplifier for 430 Mhz.

144 Its Operation SatelLife and HealthSat: The System and Dieter Klein, PhD Technical Director, SatelLife

I. INTRODUCTION organization is an international not-for-profit SatelLife to serve global uses low-earth-orbit (LEO) micro-satellites which SatelLife's mission health communication and information needs. and the exchange of information in is to improve communication A of public health, medicine, and the environment. the fields where access to emphasis is placed on areas of the world special poor technical communication and information is limited by infrastructure, economic conditions, or disasters. system SatelLife administers HealthSet, a telecommunications the exchange of information among health designed to facilitate their developing countries and to link them with professionals in satellite colleagues abroad. HealthSet uses a store-and-forward electronic mail messages called HealthSat-1 to pick up and deliver Because HealthSat doesn't depend on international worldwide. by congested telecommunications links, it is not affected service charges, or frequently disrupted circuits, unaffordable messages ground station required to send and receive service. The transceiver HealthSat comprises a personal computer, a radio from antennas. (transmitter and rebeiver) and a set of two initial In the Fall of 1991, SatelLife began its project with the installation of five experimental demonstration Tanzania, stations in Eastern Africa - Kenya, Mozambique, ground gateway-at and Zambia and the establishment of a Canadian Uganda This paper Memorial University in St. John's, Newfoundland. the technology used for HealthSat communication, describes of operation, reports on some experiences from the first months the performance of and presents thoughts and ideas for improving the current system.

II. THE HEALTHSAT SYSTEM satellite, HealthSet is currently supported by a single LEO and built by UoSAT-3 (UoSAT-14 in Amsat terminology), designed England, and Surrey Satellite Technology (SST) in Guildford, circles the earth. launched by Ariane Space in 1990. UoSAT-3 satellite is in a about every 100 minutes in a polar orbit. The of the sun-synchronous orbit which means that all passes day: late in satellite occur at approximately the same time of

145 e morning from North to South, and early in the night, from Ruth to North. The footprint of the satellite is the area on e earth from which the satellite is visible at any one time. with a diameter of Le footprint of the satellite covers on area out 6000 km.

Fig: 1: HealthSat Visibility

All transactions from the ground to the satellite must occur hile the ground station is within the footprint of the atellite. All stations within the footprint compete for the esources of the satellite. The time a ground station is within he footprint averages about 12 minutes, with less time for asses with low elevation and about 15 minutes when the atellite passes directly overhead (see Fig. 1). For users close two o the equator, the satellite will be visible typically for asses in the morning, one arising in the Northeast and one'in he Northwest, and two passes at night, one arising in the outheast and one in the Southwest. For users at the poles all asses will be visible.

I.1 Description of HealthNet Ground Station of Until now, all SatelLife supplied groundstations consist

a Kenwood TS-790 UHF/VFH Transceiver, capable of 45 Watts of output power, a EACComm Terminal Node Controller (TNC), which implements the AX-25 protocol and includes a 9600 BAUD modem linking it to the ground station computer's serial port,

146 12 13 14 15 16 17 18 Hour: 0 1 2 3 4 5 6 7 8 9 10 11 07/25/92 07/26/92 07/27/92 07/28/92 07/29/92 07/30/92 07/31/92 08/01/92 NORTH POLE 10 11 12 13 14 19 Hour: 21 22 23 24 1 2 3 ... 7 8 9 07/25/92 ------**---**---**--- 07/26/92 --*---**---**---- *------*----**---*----- 07/27/92 ----- *----*----*------*---**---**----*- 07/28/92 ------**---**---*--- 07/29/92 --**---**---**------.. ------*----*----*----

------**---**---**----- 07/30/92 -*----*---**----*----- 07/31/92 ------*----*----*-- 08/01/92 ------**---**---**--- BOSTON, MA (USA)

18 19 20 21 22 23 24 Hour: 6 7 8 9 10 11 12 13 14 15 16 17 07/25/92 ------*----*------07/26/92 ------**---**------*----**---*------07/27/92 **---**--- 07/28/92 ------**---**------07/29/92 ------**---**------07/30/92 ------*----*----*------**---**- 07/31/92 ------**---**------*----**-- 08/01/92 ------*----*------**---- EQUATOR

Fig. 2: Sample Satellite Schedules For UoSAT-3 ( * UoSAT-3 is visible )

a Kansas City Tracker Board providing frequency correction and antenna tracking, a set of omni-directional antennas (called egg beaters because of their shape; steerable antennas have been installed only at the Canadian gateway station), the ground station computer can be any fully IBM- compatible computer, including a math coprocessor and a large hard disk (currently SatelLife uses NEC 386SX/16I Power Mate computers with 2 MB memory, math coprocessor, 60 MB hard drive and a 2400 Baud modem).

147 The software used on the computer consists of: satellite tracking software (GrafTrak and Silicon Ephemeris to predict future satellite locations), Doppler Kansas City Tracker to provide real time shift frequency correction, and ground station software, a suite of programs provided by SST to manage the store-and-forward message system: PB to manage message downloading, PG to manage message uploading and PFHADD and PHS to implement the FTLO message protocol, prepare messages for uploading and extract messages after downloading).

:1.2 Description Communication with HealthSat space craft operating system allows full implementation The packet the Amateur Radio version of CCITT's X.25 if AX.25, message exchanges ;witching protocol [1]. All information and exchanges of AX.25 between ground stations and the satellite are (HDLC) frames as described in Fig. 3.

FLAG FLAG ADDRESS CONTROL PID & DATA FCS

and FLAG (hex 7F: 0111 1111) indicates the beginning a frame; end of satellite ADDRESS represents the call signs of the and the ground station, e.g. UOSAT5-11>abcde-11; (-11: broadcast mode; -12: connected mode) CONTROL identifies the type of frame: Information, Supervisory, Unnumbered, Unnumbered Information as well as a frame sequence number; to indicate PID Protocol Identifier field is used the use of a higher level protocol layer (typically hex F0: No layer 3 implemented); transmitted DATA is actual user data or other data purposes; for information sender check sequence is calculated by the FCS frame a frame and receiver and if different indicates corrupted during transmission.

~~Fig. 3: HDLC Frame Structure~~

148 In AX.25, correct receipt of all frames is to be acknowledged by the recipient to the sender, except that such acknowledgement may by choice be required only after receipt of a certain number (1 to 7) of consecutive frames. If acknowledgement is not received, the transmission is to be repeated by the sender until acknowledgement is received or a specified time has elapsed after which the connection is terminated. In contrast, other frames and particularly frames are not acknowledged.

At the ground station, all AX.25 protocol requirements including FCS are fully implemented in the TNC. In any communication with the satellite, the TNC first performs the frame check function (FCS) on every frame:

on reception of a frame the TNC accepts only frames which match the FCS and have therefore been received without error. All frames which do not match the FCS are discarded; on transmission of a frame the TNC calculates the FCS and incorporates it in the frame to be transmitted.

~~However, communication between ground stations and the satellite occurs currently in two modes:~~

connected mode which is used for uploading of messages (PG) and uses the full AX.25 protocol, and broadcast mode which is used for downloading of messages (PB) and uses only the HDLC frame structure but none of the acknowledgement procedures of AX.25 (see below).

In connected mode, all user data is carried in frames. The receiving agent, TNC or satellite, acknowledges receipt all frames to the sending agent. These acknowledgements, combined with frame sequence numbers, guaranty that all frames are received in proper sequence without any loss of data.

In contrast, in broadcast (un-connected) mode, all user information is carried in frames. The receiving agent, now only the TNC of the ground station, still accepts only correctly received frames. However, it will not acknowledge receipt to the sending agent, the satellite. Instead, the receiving station's software (PB) must keep track of received frames and must ascertain complete reception of all transmitted frames. To allow the software to accomplish this, SST defined and uses the File Transfer Level 0 (FTLO) Pacsat Protocol [3]. This protocol makes each frame into an independent "datagram" which carries with it all the information necessary to integrate it correctly in the transmitted data structure such as a file or message. Data frames missing from the structure are recorded in "hole" files and sent to the receiver for retransmission. The additional information carried is the name of the file transmitted and the

~~149 location (offset) in the file where the transmitted data is to inserted.~~

The advantage of use of connected mode is the reduction in protocol overhead for transmitting frames and therefore greater data throughput. The disadvantage is the added overhead for immediate frame acknowledgement, particularly in a noisy reception environment. In an environment such as packet radio/satellite, where interference may corrupt many frame transmissions, frequent retransmissions are required. In addition, in situations where the same information has been requested by several recipients within the same footprint, connected mode requires complete multiple transmission of the information to each and every individual recipient. Primarily the latter requirement leads to the implementation of broadcast mode. In broadcast mode, all ground stations within the footprint of the satellite can listen to and receive all frames broadcast by the satellite. If reception had been completely uninterfered, all receivers could have received the information with only one transmission. Because of the inclusion of recipient addresses in AX.25 frames, decoding of messages by unintended recipients can be prevented within the receiver software (PB). Due to frequent random interference, no two ground stations will normally receive or miss exactly the same frames. Thus, each stations software must keep its own record of frames missed (the hole list files *.HOL) and request separately retransmission of any missed frames.

~~11.3 Ground Station Operation~~

~~Ground station operation requires execution of the following steps prior to every satellite pass or every set of passes:~~

1. Program SEPH (GrafTrak and Silicon Ephemeris): Calculates satellite arrival times and frequency - corrections from orbital constants (Keplerian elements) of the satellite for the next pass or next set of passes. SEPH is an interactive program requiring operator input to some twenty questions.

~~2. Program LEPH (Kansas City Tracker): Loads arrival times and frequency-corrections into a table for use by the Kansas City Tracker antenna tracking and frequency adjusting programs.~~

3. Program PFHADD (SST): For each message to be uploaded, PFHADD adds the FTLO file header to the message. PFHADD is an interactive program requiring operator input to five questions. to proper 4. DOS COPY commands: Move prepared messages disk directory for uploading.

(SST): Start program PG for uploading of 5. Programs..PG/PB PB messages after rise of next satellite. Start program be after completion of program PG (sequencing can automatic through use of a batch file). from 6. Program PHS (SST): Strips FTLO file header downloaded messages. or 7. Print and deliver messages received by hand electronically by mail.

III. EVALUATION OF EBPERIYJMAL GROUND STATION OPERATIONS to First experiences with ground station operation proved at a exciting, often satisfying, but also often frustrating be to number of stations. Many of the problems reported pointed the problems with or inability to receive and communicate with related satellite, both on the uplink and the downlink. Several first causes have been identified and resolved. However, these have also pointed to renewed examination of the major experiences are of the whole system. Some of the causes identified components most discussed below with "quick fix solutions". However, for system problems the real solutions will come with changes to the which are being explored at this time (see section on Future Developments below).

The satellite moves so fast that transmission and reception frequencies must be corrected for Doppler shift (remember the tone of the siren from an approaching and receding police car). Calculation of these frequencies requires accurate timing in relation to Universal Time. Deviations of even twenty seconds, depending on the state of the rest of the system, may make reception marginal if not impossible.

~~Quick fix solution: compare computer clock often with on- the-hour time ticks of BBC World Service. In a future version of PB, accurate time ticks will be derived directly from the satellite clock.~~

The orbit of all LEO micro-satellites is not stable and changes continuously over time. The orbit parameters are reported in the Keplerian Element set for the satellite and are used by program SEPH. NASA updates these element sets periodically, based on actual measurements of the orbits of satellites, about every two weeks for UOSAT-3. Keplerian Elements out-of-date by a few weeks may again make reception marginal if not impossible.

~~151 Quick fix solution: the most recent set of Keplerian Elements are now always carried on the satellite and should be used as soon as available.~~

A critical component of the ground station are the antennas used for transmission and reception. As transmission power is controllable from the ground, transmission efficiency can be improved by more power from the ground. in contrast, reception is completely limited by the power received through the antenna. Early on, SatelLife made the decision that it wanted ground stations with simple, non-directional antennas. This choice significantly limits the signal gain obtained as compared with steerable antennas (as used by Volunteers in Technical Assistance) which require more maintenance and are more susceptible to breakdown.

The major problem occurs during connected mode uplinking (PG), when the sending ground station does not hear the satellite's acknowledgements, leading to frequent retransmission of acknowledgements and subsequent complete disconnects of the link.

~~Quick fix solution: none, except to keep antennas as efficient as possible (short antenna leads, good noise- free antenna locations).~~

~~At this time there do not exist any good diagnostic tools which allow station operators to assess easily and correctly the health of a ground station.~~

~~Quick fix solution: none; future releases of PB will increasingly include diagnostic tools.~~

The Kansas City Tracker employs two TSR (terminal stay resident) programs to calculate Doppler shift corrections (and antenna control with steerable antennas) in real time. These programs can interfere with various other ground station programs leading to computer freeze ups.

~~Quick fix solution: may be tricky, but may also require substitution of existing programs (see section on Future Developments below).~~

The UoSat-3 spacecraft is currently providing service in three different operating modes: broadcast mode for 'downloading (SatelLife), connected mode for downloading (VITA), and connected mode for uploading (SatelLife and VITA). In addition, certain special applications use a digipeating mode. Consequences of these mixed mode operations on overall mission performance needs to be examined in detail.

~~152 Quick fix solution: none. and Summary Lesson One: Murphy's Law holds! Trust it design a system with Murphy's Law in mind!~~

IV. LINK ERRORS AND PROTOCOLS has a The design decision to use omni-directional antennas impact on the performance capability of the whole significant to allow system. While AX.25 and its parent X.25 were designed in the error free communication, the expected bit error rates of transmission channels were assumed to be in the order highly magnitude of one bit error in ten million bits, i.e. these error reliable links. In the packet satellite situation, as radio rates may be significantly higher due to such factors interference, spacecraft wobble, atmospheric conditions, terrestrial location and - antenna performance. Current with the omni-directional egg-beater antennas experience lower indicates that the frame acceptance rate is significantly as compared to frame acceptance with steerable antennas.

SatelLife is currently engaged in various experiments to test other omni-directional, non-steerable antenna designs. Anyone interested to participate in these experiments is invited to contact the author at SatelLife.

Another approach is to examine the proper choices for the such parameters available to tune the AX.25 protocol operations, of frame size, length of time for acknowledgements and choice as add protocol all together. For example there is consideration to a parameter to the station call sign to indicate station quality. Such a parameter would allow the spacecraft to change communications parameters including transmitter power according to ground requirements.

SST is currently investigating use of broadcast mode also in the uplinking of messages, eliminating the use of frame acknowledgements all together and replacing it with hole lists, similar to their use for down linking. Use of a broadcast protocol in the uplink would eliminate the possibility of link time-outs, but would not necessarily improve the throughput capacity in itself.

~~V. FUTURE AND NOT SO FUTURE DEVELOPMENTS~~

A major goal of SatelLife is to use the simplest, most robust and durable technology available at reasonable cost. Examination of both the hardware and software used in the current station, points to the possibility for significant improvements.

~~153 V.1 Radio Redesign~~

SatelLife is currently examining various designs for a design single channel transmitter and receiver. A successful will include automatic frequency control which will eliminate the need for accurate time and Keplerian elements. completely the Reception simply starts, whenever the satellite comes over Very promising radio designs are at the horizon. horizon. ground Introduction of such radios would reduce both the cost of as well as the need for operator maintenance. In stations a components of such a station can all be operated from addition, the complete 12V power supply such as a car battery, insulating portable. station from power failures as well as making it truly

~~Summary Lesson Two: KISS - Keep It Simple Stupid (the between name of the protocol used in broadcast mode (PB) the TNC and the computer).~~

V.2 Future Ground Station And System operation current ground station operation is completely manual, While of is on the horizon. A serious problem with use automation a equipment and software has been the inability to build amateur Kenwood unified system in both hardware and software (e.g. the needs frequency correction, which requires a special transceiver and board, which requires daily updates through SEPH computer above will LEPH). Introduction of radio transceiver as describe eliminate all of the above and reduce cost. ground station are only linked to the HealthSat (see Current special Fig. 4). Any communication outside this link requires operator interference.

~~Fig. 4: Current Ground Station Design~~

~~HealthNet serves to fulfill two missions: communication in the field of health between individuals and groups in the form of electronic mail and~~

154 distribution and exchange of health information of general interest. The current design plans include exclusive use of FIDO mail outside the Canadian and European (in the future) gateway stations.

Every ground station will act as a FIDO node, which can be accessed on the ground via telephone and modems for delivery and pick-up of messages. Prior to each satellite pass, messages will be bundled by ground station destination and prepared for satellite uplink. Any mail destined for addressees on other systems will be gatewayed into the INTERNET at the Canadian station and vice versa.

Every ground station will also house an electronic bulletin board which again will be accessible on the ground via telephone and modems. The bulletin board will serve as depository for HealthNet information of general interest. Thus the next generation of ground stations will be more like those depicted in Fig. 5.

~~Fig: 5: Future Ground Station Design~~

~~VI. REFERENCES~~

~~[i] Fox, Terry L., "AX.25 Level 2 Protocol", Proceedings - 2nd ARRL Amateur Radio Computer Networking Conference, 1983 pp 4-14.~~

~~[2] "Operating Manual for PacComm Packet Controllers", PacComm Packet Radio Systems, Inc., Tampa FL.~~

~~155 "PACSAT (3) Price, Harold (NK6K) and Jeff Ward (GO/K8KA), Broadcast Protocol", UoSAT Unit, University of Surrey, Guildford, Surrey GU2 %XH, UK.~~

~~Dieter Klein, PhD Technical Director SatelLife, Inc. Tel: (617) 868-8522 126 Rogers Street Fax: (617) 868-6647 Cambridge, MA 02142 E-Mail: [email protected] USA~~

~~156 Regulatory Issues~~

~~157 158 KENYAN PACSAT LICENSING EXPERIENCE~~

~~A.J. Rodrigues & G.M. Macharia Institute of Computer Science, University of Nairobi, Nairobi, Kenya.~~

1. INTRODUCTION As part of the ESANET (Eastern and Southern African Network) Project involves experiencing in Packet Radio Satellite Communications Systems, it was agreed that=each node make its own application to their respective PTAs. for licences. This paper documents the stages involved in the process.

2. INITIAL REQUEST The following comprises the gist of the initial application made by the Deputy Vice-Chancellor (Admin. & Finance) University of Nairobi to the Managing Director, Kenya Posts & Telecommunications Co. Ltd. (KP&TC) on 21st December 1990.

The Eastern and Southern African Network (ESANET) Project, initiated by the Institute of Computer Science (ICS), Nairobi and funded by the IDRC of Canada, is to support the Computer Centres, or Institutes, of the University of Nairobi, Makerere University, University of Dar-es-Salaam, University of Zambia and the University of Zimbabwe to experiment with and evaluate various microcomputer-based data communications networking systems within the region. This communications research for development activity will be pursued collaboratively with several other organizations interested in the telecommunications and informatics areas.

The purpose of this letter is to request permission to experiment with packet radio communication using low earth orbit satellite technology as one of two strategies. This will enable us to pioneer Research and Development (R&D) in the region in Telematics and in so doing strengthen communication, between researchers and other user groups, of non-classified information within the bounds of respective national security interests.

~~3. FIRST RESPONSE FROM KP&TC The University received the first written response from KP&TC dated 30th-January 1992 which is given below:~~

~~The following is our response on the proposed Eastern and Southern Africa Network (ESANET) project:~~

~~1. The computer linkage through packet radio data communication shall be a private data network.~~

~~2. The technical and general information provided at the moment is not adequate. The University of Nairobi is~~

~~159 required to provide more detailed technical information, including at least the following:-~~

2.1 Earth Station Equipment 2.1.1. VHF packet radio: (a) VHF band giving the exact frequencies intended to be used. (b) VHF transmitter(s) power level(s) (c) VHF antenna type to be used, giving information on its radiation characteristics etc. (d) Detailed technical information of the VHF packet radio system including circuit diagrams, Block diagrams etc. (e) VHF receives technical details including receive frequencies. 2.1.2 Data Modems (a) Type, mode, make etc of the modems included for use. (b) Modems technical details including transmission speeds, modulation techniques etc. 2.2. SATELLITE SYSTEMS: Information of the satellite to be used should be. (a) Communication system details description. (b) Tracking, control and monitoring systems and site location. (c) List of other users of the satellite system apart from the said Universities. (d) Technical specifications, including equipment manuals if available.

3. Clarification is required'on how the two telephone lines to the public network are intended to be used since ESANET would be a private closed network. 4. Provide KP&TC with a copy of the list of modem types approved and those whose approval is being processed in each user country. 5. It should be noted that telecommunication equipment to be used will require type approval by KP&TC if they are to be interwork with our network system. You will also require to apply for frequency allocation after the above information is received and accepted.

~~4. CLARIFICATION FROM UNIVERSITY OF NAIROBI On the 21st March 1991 after studying carefully the letter from the KP&TC the University replied as follows:-~~

Please find attached a bound volume comprising the following documents: - Response to comments and queries raised by KP&TC - Appendix I ESANET PROJECT DOCUMENT - Appendix II WORLD-WIDE TOPOLOGY: ASSOCIATION OF PROGRESSIVE COMMUNICATIONS

160 - Appendix IV PACKET RADIO SYSTEM BLOCK DIAGRAM - Appendix V PACKET RADIO SYSTEM BLOCK OPERATION MANUAL - Appendix VI GRAFTRAK II AND SILICON EPHEMERIS SOFTWARE MANUAL - Appendix VII KANSAS CITY TRACKER SOFTWARE MANUAL - Appendix VIII AN OVERVIEW OF THE PACSAT PROTOCOL SUITE We hope this clarifies and answers the genera7 and technical points raised in your letter under reference, in sufficient depth. It is emphasized that ESANET is not intended to be private data network, in fact one entire experimentation strategy uses dial-up modems and the public network, while only the Satellite/ground-station component of the other experimentation strategy is private, other linkages being public. The latter strategy, however, as explained in the point by point response attached, is primarily to facilitate Medical Research and its usage, the users and the data volumes available, are restricted, are not-for-profit and are non commercial. Hopefully the development impact in the field of Health Care and communications will be weighed positively against the perceived loss of revenue for KP&TC once these issues are viewed in the perspective outlined here, and in greater detail in the accompanying documentation. If any further information is required we shall be only too glad to provide it.

~~The detailed response is appended.~~

~~5. SECOND RESPONSE FROM KP&TC On the 3rd May 1991, the University received the following response from KP&TC:-~~

~~The following is our response after perusal of the Esanet documents:~~

~~1. PUBLIC VOICE TELEPHONE LINES~~

1. 1 Since the two lines are intended for direct d i a l -up data communication via KP&TC network, the Univergity of Nairobi will require KP&TC's type approval of the two modems (GVC supermodem and Telebit trailblazer). As no technical information on the modems has been provided, it will be necessary to forward the manuals along with the application for type approval. 1 . 2 The -University of Nairobi w i l l apply to our telephone sales office for provision of the two telephone lines once the use of the modems is type approved. 1.3 It is noted that KP&TC may not guarantee satisfactory working of the telebit trailblazer modem due to the high speed. Such high speed modems require digitalised network for satisfactory performance.

161 2. PACKET RADIO SA TEL L I TE 2.1 Esanet being a research project and. using UOSA T-F Satellite methodology which is different from the conventional Geostationary orbit Commercial communication Satellite, the University of Nairobi would require the Kenya Government authority before KP&TC would endorse the use of the Packet Radio satellite Earth station. 2.2 The VHF/UHF up-and-down link frequencies for the packet radio equipment would have to be approved /allocated by KP& TC's radio control and licensing Section. A formal request for allocation/approval of these frequencies is necessary.

It is hoped that this response will provide the necessary guidance on the steps to be taken by the University of Nairobi to realise the Esanet project. Should you require further information or clarifications, do not hesitate to contact us.

6. REQUEST TO GOVERNMENT OF KENYA Further to the advise from KP&TC, the Vice-Chancellor, University of Nairobi made the following request to the Office of the President on 26th August 1991. The University of Nairobi is undertaking the above research project funded by the International Development and Research Centre, to experiment with various microcomputer-based communication strategies. It is a regional project involving Institutes of Computer Science/Computer Centres in five Universities in the Preferential Trade Agreement Countries, namely Kenya, Tanzania, Uganda, Zambia and Zimbabwe.

~~whereas one communication media is the public telephone network, the other methodology chosen for experimentation involves use of low-orbit Packet Radio Satellite technology.~~

An Executive Brief attached contains information on the project objectives, methodologies, media systems, security, monitoring, approved and potential users, usage purpose, restrictions and conditions. The potential benefits from the research and development perspective are difficult to ignore bearing in mind that such technologies are now available through judicious funding.

The University has formally approached the Kenya Posts and Telecommunications Corporation and we have had exchanges on various technical, management and security issues. Attached is the final letter from KP& TC dated 3rd May 1991 and your attention is drawn to section 2.1 in particular.

~~162 Accordingly, we are requesting authorization from the Government for use of the Packet Radio Satellite system by the University of Nairobi for the purpose of this research project.~~

7. AUTHORIZATION FROM OFFICE OF THE PRESIDENT The Vice-Chancellor received the following authorization from the Permanent Secretary, Provincial and Administration and Internal Security, Office of the President vide a letter dated 24th October, 1991.

Reference is made to your letter of 26th August, 1991 regarding your request that the University be authorised to use Packet Radio Satellite System for research project in conjunction with the Institute of Computer Science/Computer Centres in Tanzania, Uganda, Zambia and Zimbabwe. I have noted that the project is being funded by the International Development Research Centre of Canada and it will facilitate among others medical research. Consultations have been made and it has been agreed that the University of Nairobi should use the Packet Radio Satellite System as requested.

~~8. REQUEST TO KP&TC FROM LICENSING & FREQUENCING ALLOCATION~~

~~A request to KP&TC for allocation of frequency and licenses was made on 12th November, 1991 by University of Nairobi as follows:-~~

Point 2.1 of your letter of 3rd May 1991 advised the University of Nairobi to seek Kenya Government authority before KP&TC would endorse the use of Packet Radio Satellite. This was done vide our letter dated 26th August 1991 copy attached. Please find attached a copy of a letter dated 24th October 1991 reference No. OP. 26510011Vo1.V. from the PS, Provincial Administration and Internal Security, Office of the President. This letter authorises the University of Nairobi to use the Packet Radio System for the ESANET Research Project in conjunction with the Institutes of Computer Science/Computer Centres in Tanzania, Uganda, Zambia and Zimbabwe.

Accordingly; 1. The University requests KP&TC's endorsement of the use of a Packet Radio Satellite earth station. 2. The University requests that the following link frequencies be approved, a'nd allocated by KP&TC's Radio control and licensing section:

~~Primary VHF Up-Link Frequency 148.260 MHz Secondary VHF Up-Link Frequency 148.560 MHz~~

~~Primary UHF Down-Link Frequency 428.010 MHz Secondary UHF Down-Link Frequency 429.985 MHz~~

~~163 Frequency Tolerance: within 0.005% of the above.~~

~~9. APPROVAL FROM KP&TC The following approval was received from KP&TC on 11th June 1992.~~

We are pleased to inform you that approval has been granted to the University of Nairobi to use packet radio satellite system for the ESANET Research Project under the following conditions:- 1. (a) The Ground Radio equipment shall be operated at the frequencies stated in your letter above. (b) The radio system shall operate on "Secondary" basis i.e. operation of the system sha11 cease should it cause any interference to other existing services, but may suffer interference from existing services. 2. The radio system shall be used for research purpose only and on non-commercial basis. 3. (a) A non-refundable type approval fee of Kenya Shillings two thousand nine hundred seventy (KShs. 2,970) payable once; and a renewable annual license fee of Kenya Shillings six thousand three hundred and forty four (KShs. 6,344) will be charged for the operation of the radio system. These charges are payable in advance of putting the project into operation.

(b) Should the University of Nairobi wish to use the ESANET Project dial-up data modems to work into KP&TC's public switched network, the modems must be presented together with their technical manuals for type approval by KP&TC. A separate charge will be levied accordingly. The above conditions are mandatory and we would like you to confirm your acceptance of the same in writing.

~~It is suggested that you nominate an officer who will discuss and finalise any outstanding issues on this case.~~

10. CONCLUSION: This paper documents the license application process for the installation and use of a PASCAT ground station by the University of Nairobi. We hope this paper will assist the embarking of involved in similar activities.

164 nFVLL ENT AN SOMHUNICATIONS RSEARCH fQR REQUEST FOR CLARIFICATION REPONSE IQ KP6 ' QO ENTS AND I E RTHER JECHNICAL INFORMATION. comments and queries. ** Italics denote KP&TC's verbatim through packet radio data communication 1. The computer linkages shall be a private data network. of setting up a private data The ESANET project has no intention network. document (appendix I) ESANET is a As stated in the project microcomputer aimed at investigating various research project of the five for communications. The countries based methodologies the Preferential institutions are all members of participating are the Institutes of Trade Agreement (PTA) region. These

daHarare., Computer D Nairobi Thedgenee alogoalsi of Centres at Dar es Salaam, Lusaka a the project are to:- necbased communication a. to with microcomputer technological networking min torder to acquirep then essary capacity, and, communication b. to promote more do within the resea chf ccot ve a ffi region the specific objectives Within the scope of these general goals, of the project are :- techniques with alternative modalities and i) to experiment five nodes communications among and within the for data and Harare). (Nairobi, Dar es Salaam, Lusaka, Makerere economic, sociological and ii) to evaluate the technical, network management aspects of the communication experiments; research community within iii) to disseminate information to the and the results of the the region about the development of with a view to increasing the awareness project applications, and possibilities, stimulating new and wider and inviting feed-back on related topics; community (users make recommendations to the research iv) to authorities in the institutions) and telecommunications and modalities, and on cost effective data communication region for specific appropriate network models and policies environments and applications. emphasis is on From the above it may be noted that the Thus it is clear that microcomputer based communications systems.

165 the project has no intention of offering private data communication network services in competition to any commercial services already offered by the respective Posts and Telecommunication corporations in the participating countries.

.2. The technical and general information provided at the moment is not adequate. The University of Nairobi is required to provide more detailed technical information. Two methodologies have been proposed based on different communications media to facilitate research and experimentation:- with type- i) Public voice telephone lines in conjunction approved data modems and a microcomputer-based communications software package called FrontDoor developed (APC). by the Association for Progressive Communications Kenya has been proposed as a node in the international APC network. The topology of the world-wide APC network, with planned extensions, is shown in appendix II.

The ESANET network is in the public domain and in addition to the five Institutes/Centres of Computer Science in the region, the users are those connected to the network with the appropriate public domain technology viz. type-approved modems, etc. The information communicated is expected to be similar to that in the public network. though ii) Packet Radio Satellite. Use of this methodology, private, involves microcomputer based communications software packages, a terminal node controller (TNC), and a packet radio UHF receiver and VHF transmitter set at each 'ground' station for communication with a low-earth orbit data communication satellite using either an omni- directional whip antenna or a uni-directional Yagi-type not antenna. It must be emphasised that this facility does offer on-line communications capability; it is based on the store and forward principle (following the bulletin-board modus operandi).

~~The communication packages involved are:- path GrafTrak II for real-time display of the simulated and the coverage area of a selected satellite relative to a fixed observer,~~

~~Silicon Ephemeris, a satellite prediction program working in conjunction with GrafTrak II,~~

~~Kansas City Tracker which, using a data-base generated of the by Silicon Ephemeris, performs fine-tuning transceiver and, where applicable, adjusts the direction of the uni-directional antenna.~~

~~PACSAT ground station software.suite which includes the pb.exe module for capturing broadcast messages~~

166 for ground from the satellite and the pg.exe module The latter include compiling the station functions. down-load up-link messages and un-packing the messages. by SatelLife technology has been made available This organisation whose is a non-profit, which in developing goals are to facilitate Medical research buys time on satellites, the countries. This organisation of being UOSAT-3 (owned by the University current one Volunteers in and the capacity is managed by Surrey) body involved Technical Assistance (VITA), a not-for-profit countries. From May in technology transfer to developing purchase time on UOSAT-F, another 1991, SatelLife will payload of Surrey satellite. Owing to University and down-linked restrictions, the total volume of up-linked ground station is limited data traffic available for each station for not only the Medical to 500 pages per month per and but also the participating Institutes Faculties, Universities in the Centres of Computer Science from region. by the to the medical information exchanged In addition exchanged by the Medical Faculties, the information Institutes and Centres will be non- Computer Science will be not-for-profit and non-classified, and commercial, and project strictly limited to experimental data documentation. will be information for up-loading to the satellite Such public network received by the 'ground' station via the outlined above in 2. i). Similarly, down-linked strategy will be which is received in broadcast mode, information, same public passed on to the destination by use of the a convenient and network modality. This therefore affords the system for effective means to monitor the use of security purposes. that the only private component of From the above, it is clear As is the transceiver-satellite linkage. the second strategy the users given the strict restrictions on the,usage, such, and issue of ESANET the data volLIres as outlined above, the and in its proper providing a private data network is placed perspective.

~~equipment Below we give detailed technical information izoutthe involved in the second methodology v Packet Radio Communication using low-earth orbit data satellite.~~

~~2.1. Earth Station~~

~~2.1.1 HF VHF Packet Radio~~

~~Kenwood Transceiver Model TS-790A~~

167 For detailed specifications see page 4 and 5 of the Kenwood Transceiver Instruction Manual in appendix III. intended to be (a) UHF/VHF band giving the exact frequencies used:- Primary VHF Up-Link Frequency 148.260 MHz MHz Secondary VHF Up-Link Frequency 148.560 MHz Primary UHF Down-Link Frequency 428.010 MHz Secondary UHF Down-Link Frequency 429.985 above Frequency Tolerance within 0.005 X of the

~~(b) VHF transmitter power level~~

Basic Up-Link transmitter power level in SSB Operation:- Maximum 35 Watts Authorised (EiRP) 10 Watts Modulation Up-Link Emission Type is Narrow Band Frequency level in NBFM i.e. NBFM (F3E). Maximum VHF Transmitter power mode is:- 16.5 dBW, Max. on (c) VHFIU14F antenna type to be used, giving information its radiation characteristics etc.

~~The VHF Up-Link antenna used is either: i) with: VHF Omni-Directional Turnstile Reflector (TR) array~~

~~Gain: 2.5 dBi 3 dB BW: 140 deg. (E & H) Polarisation: RHCP Resultant EiRB: 19.0 dBW EiRS~~

~~OR of two planes. VHF Yagi-Uda Array with 9 elements in each~~

Gain: 13.2 dBi (H) 3 dB BW: 19.5 deg. (E) / 22.2 deg. Polarisation: RHCP Resultant EiRP: 29.7 dBW EiRP Max. radio (d) Detailed technical information of the VHF packet system including circuit diagrams, Block diagrams etc. appendix Iv. See the overall System Block Diagram in Kenwood Transceiver manual in appendix III. See the and the details of the Terminal Node Controller(TNC) For manual for Modems, see 2.1.2 below and the operation Data appendix V. the NB-96 Series PacComm Packet Radio System in

~~168 (e) VHF receiver technical details including receive frequencies.~~

~~See 2.1.1 (a) and (d) above.~~

~~2.1.2 Data Modems:~~

The Data Modem and the Terminal Node Controller (TNC) are both integrated into each PacComm packet radio (NB-96 model) manufactured by PacComm Packet Radio Systems Inc. of Tampa, Florida, USA. This unit is widely used by licensed radio amateurs world-wide.

In the down-link mode, the TNC receives digital data at 9600 bps from the data modem and sends it to the IBM PC compatible computer via a RS 232C link. Similarly, in the up-link mode, the TNC recieves the data file from the PC via the RS 232C link and passes it to the modem which modulates an FM signal using the FSK (Frequency Shift Key) method.

~~The TNC implements a sub-set of the CCITT's X.25 communications protocol functioning at OSI/ISO level 2.0 (i.e. at the Link Level).~~

~~Further details are in the Operations manual of the PacComm Packet Radio System in appendix V. In (a) and (b) below a few key specifics are presented.~~

~~(a) Type, mode, make etc of the modems included for use.~~

~~Make: PacComm Model: NS-96 Manufacturer: PacComm Packet Radio Systems, Tampa, Florida.~~

~~(b) Modems technical details including transmission speeds, modulation techniques etc.~~

Speed: 9600 bps Technique: FSK Modulation: Direct FM +/- 3.0 KHz with an RF spectrum of less than 20 Khz at the -60 dB points. Transmit Modulator: 8-bit long digital F.I.R transversal filter in EPROM for transmit wave generation with a 'brick wall' audio spectrum (typically -6 dB at 4.8 KHz, and -50 dB at 7.5 KHz). Receive Demodulator: Audio derived from the rf receiver discriminator. Fed into a third order Butterworth filter, 6KHz. Clock Recovery: PLL clock recovery with 1:256 resolution.

~~169 2.2. SATELLITE SYSTEMS:~~

The UOSAT spacecraft in low-earth Bun-synchronous orbit are controlled from the UOSAT Mission Control Centre at the University of Surrey, Guildford, UK. The control centre is completely automated and thus with the space-craft's on-board computers (OBC), the entire system requires minimal human intervention. A ground-based network of computers enables a wide range of digital data from the spacecraft to be received, displayed and archived on the University's Main-frame computer.

The spacecraft is gravity-gradient stabilised with magna-torquing coils under OBC control to induce and control a slow axial spin for thermal equilibrium. The on-orbit average power generated by the Gallium Arsenide (GaAs) solar arrays is 15 Watts.

Spacecraft housekeeping is under on-board computer (OBC) control. A full suite of sensors are read by the OBC and sent periodically over the main communications channel. A command decoder receives and implements spacecraft configuration changes. of Each ground station such as the one proposed at the Institute Computer Science, Nairobi, is capable of tracking the spacecraft by use of a computer program called Seph in conjunction with spacecraft flight data updates regularly received from the mission control centre. Seph is run, and the satellite data for stored, on an IBM PC compatible computer, which is also used ground-station functions (see 2. ii). above). Future command stations will be established in the USA, Canada, UK and elsewhere as required by the projected growth in operations. The command station is similar to any other Packet Radio Satellite groundstation, the distinguishing feature being the ability to execute 'tele-commands' on the spacecraft. the The following sub-sections furnish further details of satellite communication system.

~~Information of the satellite to be used should include:-~~

~~(a) Communication system details description.~~

The communications payload consists of an up-link receiver operating in the 140 MHz range, the down-link transmitter operating in the 420 MHz range, and the Packet ConYnuni'cations Module (PCM). The PCM uses to Intel 80036 class of microprocessor hardware running at 8 MHz anc supported by 16Mbyte of RAM. The PCM, which provides computational and storag( facilities to the entire communication network, recognises up-link requests from the users, controls access, store; messages and down-links messages to users. It can engage it totally automatic transactions with -round stations.

170 narrowband FM using the FSK technique at The down-link is is bps. Thus a narrow band FM receiver/demodulator 9600 this, a 40 KHz employed at the ground station. For both the doppler shift channel, including an allowance for and a guard band, is required. transmitter's output power is 10 dBW maximum. The spacecraft (7 is normally operated with a 3 dB backoff The transmitter receiver noise dBW or 5 Watts). The spacecraft's Kelvin. temperature level is approximately 125

~~Absolute maximum data transfer: 106,600 pages per day 90% of channel bandwidth than 10 E-12) Error Rate: Negligible (less~~

~~Communications Payload: Transmitter: Frequency: 420 MHz range backoff Power: 10 dBW with 3 dB Antenna: 0.0 dBi Linear EiRP: 10 dBW Max. Modulation: FSK +/- 3.5 KHz. Link Rate, Data: 9600 bps~~

~~Receiver: channels: primary, Frequency: 140 MHz range (3 secondary and command) Antenna: 0.0 dBi Linear Noise Temperature: Approx. 125 K Modulation: FSK +/- 3.5 KHz. Link Rate, Data: 9600 bps~~

and site location. (b) Tracking, control and monitoring systems packages involved in the ground station The communication and satellite for spacecraft tracking, station control, monitoring are:- display of the simulated path GrafTrak II for real-time satellite and the coverage area of a selected relative to a fixed observer, the Ephemeris (Seph),a program that calculates Silicon its epheme ris position of the satellite by knowing time of the day. (in keplerian co-ordinates) and the location of the ground The program also knows the and station expressed as latitude, longitude This information is then used both to altitude. from the the satellite viewing angle calculate Kansas City station, and to provide input to the Tracker program.

171 generated Kansas City Tracker which, using information by Silicon Ephemeris, may perform any of two functions: feed frequency correction data to the transceiver a) and the to compensate for the Doppler shift satellite movement. control for the b) where applicable, provide steering and Yagi-type directional antenna i.n the azimuth and the elevation.

includes the PACSAT ground station software suite which module for capturing broadcast messages from pb.exe for ground the satellite and the pg.exe module compiling the station functions. The latter include up-link messages and un-packing the down-load messages. i II and Silicon See the the software manuals for GrafTrak VI) and for Kansas City Tracker Ephemeris (appendix Protocol (appendix VII). See also the article "PACSAT Suite:- an Overview" (appendix VIII). this case the The ground station site location, in parameter for Institute of Computer Science, is an input and is entered as both GrafTrak II and Kansas City Tracker geographical co-ordinates, i.e. metres NAIROBI f 1.0 South, 36 East and 1600

Access Windows: However due Each access window lasts up to 18 minutes. effects, passes are only usable when the to various the horizon. is not less than 5 degrees above satellite approximately 15 Usable passes are limited to a maximum of above the 5 deg. minutes each with the actual time spent constraint varying from one pass to another. elevation passes per day. Observers on the equator see the fewest time in two usable In Nairobi, Kenya, the total access 35.82 minutes per passes above the 5 degrees elevation is day, maximum. traffic monitored It is feasible to have all ground station in section 2.i) on-line by KP&TC using the system described above. system apart from the (c) List of other users of the satellite said Universities. community of radio Other users comprise an international the five countries it amateurs, Satellife, and potentially, ESANET project subject ti the PTA region involved in the and Telecommunication: approval by their respective Posts will have the Memoria' authorities. In addition UOSAT-F University of Canada as a user primarily for down-loadin!

172 users and up- database queries from SatelLife Medical back to these users. linking Medical data for transmission is automatically monitored byUOiSAfoG dmission All traffic Surrey, control centre at the University of equipment manuals if (d) Technical specifications, including available. UOSAT-3 UOSAT-F (up to May 1991) (From May 1991) 800 Kms. Height: 810.5 Kms. 101.0775 Minutes 100.48 Minutes Period: 98.5 degrees inclination: 98.7427 degrees

Orbital Rate: 14.33 Orbits/day i Orbit of Nodes: 25.2 degrees West Per Progression degree elevation Maximum Foot-print: 6076 Kms at 0 +/- Doppler Shift: 4140 MHz up-link: 20 MHz down link: +/3.10 KHz.

lines to the is required on how the two telephone 3. Clarification ESANET would be a public network are intended to be used since private closed network. ESANET is not a As emphasised in sections 1.0 and 2.0 above, private closed data network. lines connected The two telephone lines requested are out-of-area to f acilitate:- to the main public exchange at Nairobi Central reliability and availability to i) greater telephone line domain ensure effective utilisation of the public i) above, communication strategy outlined in section'2. ground communication of information between the ii) efficient and the station, at the Institute of Computer Science, Medical Faculties, and, of tho contents effective on-line monitoring by the KP&TC iii) using the system of the up-link and down-link data files described in section 2. 1) above. these two lines will be paid from The cost for installing for the modems project funds as will any type approval fees listed below in section 4.0.

173 4. Provide KP&TC with a copy of the list of modem types approved and those whose approval is being processed in each user country. type The data modems listed below will be used exclusively, after approval, with the public domain network outlined in section 1 as stated in and 2. 1) above for the purposes of experimentation the project document, see appendix I.

a) GVC Supermodem series of internal and external modems. 2400 bps, supporting up to V.22 bis standard. rates b) Telebit Trailblazer high speed modem cap[able of data up to 19200 bps, supporting V.32 bis standard and MNP level 4.0 error correction.

~~for the above modems is currently being sought by Type approval and the four other institutions from their respective posts telecommunications authorities in the PTA region.~~

to be used 5. It should be noted that telecommunication equipment interwork will require type approval by KP&7'C if they are to be with our network system. above Due cognizance of this point has been taken as exemplified in sections 1., 2.(1), 3. and 4. the You will also require to apply for frequency allocation after above information is received and accepted. and We hope that the information provided above is adequate detailed enough. We will be glad to furnish further information that may be if necessary and to comply with any formalities required.

~~APPENDICES~~

~~1. ESANET PROJECT DOCUMENT COMMUNICATIONS II. WORLD-WIDE TOPOLOGY: ASSOCIATION OF PROGRESSIVE~~

~~III. KENWOOD TS-790A TRANSCEIVER INSTRUCTION MANUAL~~

~~IV. PACKET RADIO'SYSTEM BLOCK DIAGRAM MANUAL V. PACCOMM NB--96 PACKET RADIO SYSTEM OPERATIO14AL MANUAL VI. GRAFTRAK II AND SILICON EPHEMERIS SOFTWARE~~

~~VII. KANSAS CITY TRACKER SOFTWARE MANUAL~~

~~VIII. "PACSAT PROTOCOL SUITE:- AN OVERVIEW"~~

~~174 Regulatory Issues - Summary of discussions~~

the The workshop discussions were based on several key themes. One of these concerns of this regulatory issues which must be dealt with if the adoption, testing and utilization technology will be made possible widely. with the Participants from several countries in Africa discussed their experiences in dealing helped clarify authorities. Several of the representatives from the PTTs also provided input and where digital possible courses of action in the future. Experiences were shared from countries licensing radio communication systems were already installed and in use and those where Zambia, processes were still on-going or just completed. These included Tanzania, Kenya, Djibouti, Mozambique, Cameroon, Somalia, and Chad.

~~Regulatory issues centered around several concerns; security, competitive restrictions, frequency allocation, licensing processes, duties and excise taxation.~~

~~It was felt that, in the long run, the PTTs should continue to provide services, but policies related to the services be established by another body.~~

Frequency Allocation and Licensing for Operation of building In the presentations, it was made clear that this is often the most difficult part digital radio communications systems. The experiences showed that:

there is little or no awareness of these technologies in most countries, particularly among PTTs and governments in general; it no specific policies are in place as of yet in most of these countries, thus making difficult to allow for easy installation and operation of the systems; and

~~the type of information required before licenses could be issued was varied, generally too much was required , and often in bits and pieces over time.~~

are In all cases, it was clear that frequency allocation and licenses to operate the systems required. However, obtaining them could be a rather long and complicated process. In Kenya for example, it took 18 months from the time the first application was submitted, to the time the approval was granted. The participants concluded that the consultation process between the proposers/sponsors of the system and the authorities should be started at the earliest time possible. mutual It is critical for the sponsors to clarify that the process and undertakings are for the in benefits of both the developers/users and the PTTs. If possible, the PTTs should be included was done in the projects as an educational activity, particularly for the technical people. This Zambia and worked well.

Efforts should also be made to include members of the governments and PTTs more generally issues of the in communication project development for them to understand the "management" specific projects.

175 It is necessary to develop a broad awareness concerning information technology within governments and PTTs to facilitate the formulation of broad enabling policies, particularly as they relate to the social and economic development field. This will help greatly to change the attitude of the governments.

On the part of the NGO's and other organizations/institutions affected by the regulations, there is a need to develop more positive attitudes towards PTTs and governments when going through the process. They should also be ready to commit the resources necessary to make this possible.

In all cases, where participants have patiently gone through the system of working through the process with the PTTs and relevant government agencies, there is a great appreciation for the long term impact on policies. If short cuts are adopted, they may prove to be a liability as the "approval" may only be short lived.

It was pointed out that the allocation of frequencies by the World Administrative Radio Conference (WARC) for LEOs should make the process of licensing easier in the long run in individual countries. The availability of information on events such as the WARC decision and about countries where the licenses and applications have been achieved will also expedite the process in new countries.

~~Duties~~

The customs and excise duties levied on communication equipment in most countries and the requirement for special authorizations to import the equipment are prohibitive. Unless these taxes are waived, the costs of the equipment end up being too high and loses economic sense, particularly given the fact that the systems are meant for humanitarian and development purposes only - the process the technologies are meant to expedite. In most cases, the users of these systems have sought exemption from paying the duties.

Where necessary, donor agencies and governments should work towards the facilitation of custom waivers and reduce regulatory impediments, thereby increasing affordability and access to chosen technologies.

In conclusion, since the use of digital radio technology is being developed as non-profit, non- competitive systems for the encouragement of social/economic development, the participants generally felt that governments should ease the regulatory constraints on the establishment of such systems, including but not restricted to PTT approval processes, customs and excise regulations.

~~176 Training Issues~~

~~177 178 Training Issues Summary of Discussion~~

techni- group was quite interested in the subject of ongoing The uses to cal training regarding the technologies and potential they can be put. An additional point made was that improve- which of and changes are continuous and there should be a means ments presented keeping up with these as well. There was some evidence ways inade- that documentation as it currently exists is in some and either needs to be constantly revised and/or a'means quate basis. found for technical support to be provided on an ongoing of technical assistance as a service would not neces- Provision to use. sarily indicate that the technologies are too difficult commercial Rather, it would follow the model used now by many vendors of software in supporting their products. should At the same time, more emphasis on local problem-solving be made in the documentation that exists, so that "customer support" services become the "place of last resort." Trouble- shooting sections in documentation could be enhanced through polling current users for their input on what the common difficulties have been and solutions found to those problems.

I to The question of whether technologies should be made as simple operate as possible or whether to invest more in user training was not decisively answered. The group was clearly pre-disposed and toward making the most of existing training opportunities be creating more. It was felt that more training slots should made available within Africa, not in the United States or in region- other places outside Africa. Emphasis should be placed on based training, as opposed to national programs. This was because training programs established on a national basis might tend to discriminate in favor of the stronger African countries.

Where hardware/software cost is an important factor, consensus seemed to exist on keeping those expenses low while investing more in training. The basic "rule of thumb" is that the fewer knobs there are on the radio, the more expensive the radio is. This is because the radio is being asked to do more and the operator less. It was also pointed out that the drive to make the systems easier from a user perspective frequently comes not from the users themselves, but by non-Africans.

Besides costing more, an additional trade-off is that the easier the systems are to use, the less local adaptation is generally possible and the more users will need to depend on outside, generally non-African, consultants. The literature of technology transfer clearly shows that local adaptation and "re-invention" are usually prequisities for system vitality.

~~on the other hand, not everyone wants or needs to become a~~

179 technical expert. Users want to send and receive messages and files to support their programmatic needs without spending a lot of time on the technical details that make the communications possible. Balance is required. One possible scenario is to create and support a cadre of African technicans to provide the local technical back-stopping for a user community that depends on them and not on non-African consultants.

It was also suggested that computer communications training should be included in existing computer science curricula where it has traditionally been given short shrift (including in the United States). A course which combines computer literacy and communications training using the computer could be an important addition to the existing training milieu. There should be some means to encourage students to take digital communications training, possibly by showing that by doing so their careers and/or employment potential is enhanced. So any training should have practical aspects as well, not just theoretical formula- tions. Practical, "hands-on" training opportunities could be offered by organizations outside the formal academic institutions.

It is in this last aspect that training relates to the regulatory issues presented, because to the extent that governments encourage more groups and individuals to use the low-cost radio-based technologies, the more individuals will see that it is to their benefit to learn more about these technologies. A possible rationale to use when urging governments to relax licensing requirements is to point out that local technical competence will be created increasing the country's base of technical human capital.

~~180 Conclusions~~

~~181 182 Conclusions~~

by The International Workshop on Digital Radio Technology and Applications was organized VITA and IDRC in collaboration with UNU with the objectives of: recipients, PTT - bringing participants in current digital radio projects, prospective project representatives, and donors together; and allowing participants to acquire a practical knowledge of digital radio technologies applications;

promoting useful interaction and understanding through the exchange of experiences with existing applications; and to defining the related training, regulatory and implementation problems/issues in order facilitate appropriate utilization of these technologies in the future.

The general consensus seemed to be that it was a successful event and, to a large extent, individuals these objectives were achieved. The Workshop was very well attended - over sixty registered from twenty-nine countries. A good balance of experts, relevant project personnel, was too PTT representatives, and potential users participated. Some participants felt that there much material to cover in three days - it was too intensive. Others wanted more basic introductory coverage of the technologies. In a single event, however, it is difficult to avoid this latter situation as participants have different personal agendas and different levels of expertise and experience.

The Workshop was not planned with the objective of producing specific recommendations. However, some conclusions, both general and specific, were discernible through the course of the workshop. These are presented below.

There was general agreement that this technology was filling a niche in support of economic non- and social development at this point in history; that the establishment of non-profit, competitive systems should be encouraged. With time, PTT-supplied infrastructure, with appropriate pricing, would supplant the need for this technology in some locations (and eventually, perhaps, in all locations).

On the policy side, participants saw merit in broadening awareness of the role of information and communication technology within all sectors of the development community (government, academia, private sector, NGOs, donors). In some countries, project participants felt that it was important to work with PTT personnel, to involve them in the projects from the beginning. Other participants felt that involving PTTs in projects risked the possibility of blockages being' caused. Some of the PTT representatives felt that service and regulatory functions of PTTs should be separated. Most participants felt that regulatory constraints should be eased in order to increase affordability of and access to appropriate technologies. Concerns were raised these during the workshop about the need to make the design, implementation and operation of communication systems a more indigenous process and the need for policies which would nurture this.

183 Training was seen as a critical element in the transfer of these technologies. Focus should be placed on regional, short-term training in the South. VITA felt that USTTI might be interested in replicating their packet radio course in developing countries.

Of technical interest, the benefits of low-cost, fixed, omnidirectional antennae (implying less satellite access time) versus higher cost, steerable antennae (implying greater satellite access time) were debated. Although there was no conclusion, with 90% of a pass being accessible with steerable antennae versus approximately half that with fixed ones, it is tempting to side with the more sophisticated technology. One must counterbalance this conclusion with cost, implementation and maintenance issues. If the application requires the transfer of large files, the steerable antennae appear to be necessary. The satellite designers, Surrey Satellite Technology Ltd. (SST), were interested in the views of participants on how best to handle increased demand resulting from the expansion of the number of ground stations. The response can be more satellites, greater throughput, multiple channels, or a combination of the three. They would appreciate feedback in the future.

Many of the participants from developing countries expressed the need for additional documentation including technical literature, source books on alternative technologies, a glossary of terms, and directories of experts and training opportunities.

It was a pleasure for the organizers to see this workshop take place successfully after many years of effort promoting the technology and planning for the event. Originally the workshop would have taken place before there was serious experimentation underway, at a time when there were strong concerns that authorization for the use of this technology might be very difficult to obtain. As it transpired, the workshop took place with many licenses having been granted and extensive, albeit nascer%t, experience already accumulated. This workshop has allowed the majority of the key players to come together and share experiences at the point where many of the initiatives are about to expand their involvement and utilization of the technology. It was a timely and worthwhile event for all involved.

~~184 Appendices~~

~~185 186 Agenda~~

~~International Workshop on Digital Radio Technology and Applications - Workshop Agenda~~

~~Safari Park Hotel, Nairobi, Kenya 24-26 August 1992~~

~~Monday 24 August~~

~~Morning 08-30 - 09:30 Registration~~

~~09:30 - 10:00 Welcome VITA - H. Baiya IDRC - D. Balson VITA - G. Garriott UNU - I. Wesley-Tanaskovic~~

~~10:00 - 10:30 Official opening~~

~~Managing Director, Kenya Posts and Telecom~~

~~10:30 - 10:45 Coffee Break Appointment of 10:45 - 11:15 Workshop Objectives, Agenda, Scheduling, Logistics; Drafting Committee for Recommendations - D. Balson~~

~~11:15 - 12:30 Digital Radio Technology Overview - G. Garriott (VITA)~~

~~12:30 - 14:00 Lunch~~

~~Afternoon 14:00 - 15:45 Digital Radio Applications; Chair, D. Balson~~

~~Healthnet Zambia M. Bennett Healthnet Kenya W. Okelo-Odongo Healthnet Tanzania W. Sangiwa Tanzania/U. Dar H.R. Mgombelo~~

~~15:45 - 16:00 Coffee Break~~

~~16:00 - 17:30 Digital Radio Applications; Chair, G. Garriott~~

~~Chad/VITA T. Yondailaou Philippines J. Guillermo~~

~~187 Tuesday 25 August~~

~~Morning 08:15 - 09:30 Digital Radio Applications (Cont'd); G. Garriott~~

~~Uganda P. Mugambi Lesotho K. Sonopo Sierra Leone/PLAN Inter. J. Sandi Djibouti/ISERST D. Djama Mozambique/CARE P. Gray Pakistan J. Rushby~~

~~09:30 - 10:15 Satellite Based Development - Past, Current, Planned - Chair, S. Ramani~~

~~University of Surrey M. Sweeting PACSATNITASAT G. Garriott SatelLife D. Klein~~

~~0:15 10:30 Coffee Break~~

~~10:30 - 12:30 Satellite Demonstration - ICS, UON~~

~~Afternoon 14:00 - 15:45 Satellite Based Development (Cont'd)~~

~~15:45 - 16:00 Coffee Break~~

~~16:00 - 17:30 Regulatory Issues Panel Session - Chair, P. Mugambi~~

~~17:30 - 19:30 Break~~

~~19:30 - 21:30 Training Issues Panel Session - Chair, 1. Wesley Tanaskovic~~

~~188 Wednesday 26 August~~

~~Morning Digital Radio Networks Panel Session - Chair, 09:00 - 12:00 Designing/Implementing M. Bennett~~

~~10:30 - 10:45 Coffee Break~~

~~Afternoon and Technical Overview; P. 14:00 - 15:45 Packet Radio Technology Demonstration Gray, D. Klein, G. Garriott~~

~~Concurrent Meeting of Drafting Committee~~

~~15:45 - 16:00 Coffee Break~~

~~16:00 - 18:00 Concluding Session~~

~~189 Participant List~~

~~INTERNATIONAL WORKSHOP ON DIGITAL RADIO TECHNOLOGY AND APPLICATIONS~~

~~1. NAME: Mr. Mohammed Ahmed COUNTRY: Ethiopia INSTITUTION: Ethiopian Telcom Authority P.O. Box 1047, Addis Ababa, Ethiopia TEL: 514369 FAX: 251-1-51777~~

~~2. NAME: Professor A.J. Rodrigues COUNTRY: Kenya INSTITUTION: University of Nairobi P.O. Box 30197, Nairobi TEL: 447870 FAX: 447870 E-mail: [email protected]~~

~~3. NAME: Dr. Alex Tindimubona COUNTRY: Kenya INSTITUTION: African Academy of Sciences P.O. Box 14798, Nairobi TEL: 802182176/83 FAX: 802185 E-mail: [email protected] ([email protected])~~

~~4. NAMES: Mr.Yondailaou gue. Tolloum COUNTRY: Chad INSTITUTION: Administrator P.O. Box 1109 Ndjamena, Chad TEL: 514000/514148 FAX: 235 51 5884~~

~~5. NAME: Prof. M. Sweeting COUNTRY: UK INSTITUTION: University of Survey Guildford Survey England GU2 5XH TEL: 44 483 509143 FAX: 44 483 34139~~

~~6. NAME: Mr. Joseph Sandi COUNTRY: Sierra Leone INSTITUTION: Plan International 9 Old Signal Hill Road Congo Cross FreeTown, Sierra Leone TEL: 2304421231659 FAX: 232 30442~~

~~190 7. NAME: Mr: Lyimo H.J.R. COUNTRY: Tanzania INSTITUTION: Tanzania Post & Telecomm. P.O. Box 71524, Dar-es-salaam, Tanzania TEL: 255 51 35995 FAX: 255 51 46777 ..~~

~~8. NAME: Mr. Hanna A. Samak COUNTRY: Egypt INSTITUTION: A.R.N.T.O P.O. Box 795 Cairo, Egypt TEL: 777875 FAX: 771306~~

9. NAME: Dr. Wesley Tanaskovic Ines COUNTRY: Italy INSTITUTION: United Nations University ICTP Box 586, 34100 Trieste, Italy TEL: 39 40 224471 FAX: 39 40 224600 E-mail [email protected] stauka@itsictp (BITNET)

~~10. NAME: Mr. Mark Bennett COUNTRY: Zambia INSTITUTION: University of Zambia P.O.Box 32379 Lusaka, Zambia TEL: 260 1 252507 FAX: 260 1 253952 E-mail [email protected] [email protected]~~

11. NAME: Mr. Josefito Guillermo COUNTRY: Philippines INSTITUTION: Phil. Cooperative Banks Foundations, Inc Cooperative Banks of Davao City Lapu - Lapu St. Agdao, Davao City TEL: 642-72/771-84 FAX: 7-6-7-23 E-mail: ACDI

12. NAME: Dr. John B. Black COUNTRY: Canada INSTITUTION: University of Guelph Ottawa, Canada, N1G 2WI TEL: 519 824 4120 ext. 2181 FAX: 519 824 6931 E-mail [email protected] [email protected] JB.BLACK (ENVOY/inet)

~~191 13. NAME: Dr. W. Okelo-Odongo COUNTRY: Kenya INSTITUTION: Institute of Computer Science Univ. of Nairobi P.O. Box 30197, Nairobi TEL: 254 2 442121 FAX: 254 2 44 7870 E-mail: [email protected]~~

14. NAME: Mr. Lishan Adam COUNTRY: Ethiopia INSTITUTION: Pan African Development Information System (PADIS) P.O. Box 3001 Addis Ababa, Ethiopia P.O. Box 5834, New York, NY-10163, U.S.A. TEL: +251 1511167 FAX: 251 1 514416 E-mail: [email protected] [email protected]

~~15. NAME: Mr. Stefano Trumpy COUNTRY: Italy INSTITUTION: CNUCE Institute of CNR 36 VIA 5. Mazia 56100 PISA Italy TEL: 3950593286 E-mail: [email protected]~~

16. NAME: Mr. Luvembe Kigada COUNTRY: Kenya INSTITUTION: George Washington University 8332 Draper Lane Silver Spring MD 20910 USA TEL: 301 587 7516 FAX: E-mail: [email protected] (INTERNET) [email protected] 109.z1.fidonet.org

~~17. NAME: Drouilh Suzanne COUNTRY: UN INSTITUTION: Sudano Sahelian Office P.O. Box 30552 TEL: 2542 520289 FAX: 2542 520874 E-mail: TCN- UNDP168 DIALCOM: 41:UNDP168~~

~~18. NAME: Mr. Edmund B. Katiti COUNTRY: Uganda INSTITUTION: Uganda Computer Society P.O. Box 5950, Kampala, Uganda TEL: 256 41 257951 FAX: 256 41 245938 E-mail:~~

~~192 19. NAME: Mr. Dirieh Djama COUNTRY: Djibouti INSTITUTION: Iserst Djibouti P.O. Box 486, Djibouti Rep. of Djibouti TEL: 253 35 27 95 FAX: 35 48 12 E-mail:~~

20. NAME: Dr. Gary Garriott COUNTRY: USA INSTITUTION: Volunteers in Technical Assistance 1600 Wilson Blvd, Suite 500 Arlington, Virginia 22209 USA TEL: 703 276-800 FAX: 703 243-865 E-mail: Gary.Garriott@f 165.n109.z1.fidonet.org; [email protected] INTERNET; vita@GMUVAX BITNET: 1:109/165 Fid ne

~~21. NAME: Mr. David Balson COUNTRY: Canada INSTITUTION: I.D.R.C. P.O. Box 8500, Ottawa, Ontario, Canada K1 G 3H9 TEL: 613 236 6163 FAX: 613 238 7230 E-mail [email protected] [email protected]~~

22. NAME: Dr. Amy Auerbacher Wilson COUNTRY: USA of Science INSTITUTION: American Association for the Advancement 1333 H St. N.W, Washington, DC USA 20005 TEL: 202 326 6778 FAX: 202 289 4958 E-mail: [email protected] [email protected] TELEX: 248933 SCIEN UR

~~23. NAME: Mr. S.M. Kundishora COUNTRY: Zimbabwe INSTITUTION: University of Zimbabwe Fac. of Engineering P.O. Box MP. 167, Mount Pleasant, Karare, Zimbabwe TEL: 303211 Ext 1199~~

~~193 24. NAME: Dr. Paulos Nyirenda COUNTRY: Malawi INSTITUTION: University of Malawi Chancellor College P.O. Box 280 Zomba Malawi TEL: 265 522 222 FAX: 265 522 046~~

25. NAME: Prof. Mohammed S. Sheya COUNTRY: Tanzania INSTITUTION: Tanzania Commission for Science and Technology P.O. Box 4302, Dar es Salaam, Tanzania TEL: 0255 51 74015 FAX: 0255 51 74015 E-mail: [email protected] 5:733/1

~~26. NAME: Mr. Smith Malcolm COUNTRY: Switzerland INSTITUTION: UNHCR 154 Rue De lausanne CH 1202 Geneva TEL: + 41 22 739 8227 FAX: + 41 22 731 9546~~

~~27. NAME: Mr. A.S. Dlamini COUNTRY: Zambia INSTITUTION: PRET PREFERENTIAL TRADE AREA (PTA) P.O. Box 30051, 10101 Lusaka, Zambia TEL: 260 1 229725 FAX: 268 1 252524~~

~~28. NAME: Mr. Agbor Tabi John COUNTRY: Cameroon INSTITUTION: Ministry of Posts & Telecommunications CETCAM Dirtelcam Ministry of PTT Yaounde, Republic of Cameroon TEL: 237 22 36 30 FAX 237 23 16 63~~

~~29. NAME: Mr. Charles Musisi COUNTRY: Uganda INSTITUTION: Makerere University Uganda P.O. Box 7062, Kampala TEL: 559712 or 530024 FAX: 256 41 532440 E-Mail: [email protected]~~

~~30. NAME: Mr. Rob Borland COUNTRY: Zimbabwe INSTITUTION: University of Zimbabwe P.O. Box MP - 167 Harare, Zimbabwe TEL: 263 4 303211 ext 1492 FAX: 263 4 732828 E-mail: [email protected]~~

194 31. NAME: Ms. Regina Cammy Shakakata COUNTRY: Zambia INSTITUTION: University of Zambia Medical Library P.O. Box 50110, Lusaka, Zambia TEL: 260 1 250801 FAX: 260 1 253952 E-mail: [email protected] [email protected]

32. NAME: Mr. Bob Barad COUNTRY: USA INSTITUTION: Baobab Communications 1415 21st Street, NW 2A Washington DC 20036 TEL: 202 775 1955 BBS: 202 296 9790 E-mail: Bob. [email protected]. [email protected] FidoNet: 1:109/151

~~33. NAME: Mr. Ahmed Y. Habbane COUNTRY: Djibouti INSTITUTION: IGADD P.O. Box 2653, Djibouti, Republic of Djibouti TEL: 253 354050 FAX: 253 356284~~

~~34. NAME: Mr. H.R. Mgombelo COUNTRY: Tanzania INSTITUTION: University of Dar es Salaam P.O. Box 35131, Dar es Salaam, Tanzania TEL: + 255 51 49168 FAX: + 255 51 48602~~

~~35. NAME: Mr. K. Sonopo COUNTRY: Lesotho INSTITUTION: Lesotho Telecomms Corp. L.T.C Box 1037, Maseru TEL: 324211 FAX: 310091~~

~~36. NAME: Mr. John A. Villars COUNTRY: Ghana INSTITUTION: Ghastinet - CSIR P.O. Box M. 32, Accra, Ghana TEL: 773529 E-mail: [email protected]~~

~~37. NAME: Prof. Oguulade Davidson COUNTRY: Sierra Leone INSTITUTION: University of Sierra Leone URDS, University of Sierra Leone, PMB Freetown TEL: 232 21 226325 FAX 232 22 224439~~

195 38. NAME: Mr. Nobile Alvise COUNTRY: Italy INSTITUTION: International Centre for Theoretical Physics NBIL I.C.T.P Stt. Costierd II - 34100 Trieste TEL: +39 40 2240391 FAX: +39 40 224163 E-mail [email protected]

39. NAME: Dr. Dieter Klein, Technical Director COUNTRY: USA INSTITUTION: Satellife USA 126 Rogers St., Cambridge, MA 02142 TEL: 617 868 8522 FAX: 617 868 6647 E-mail: [email protected] [email protected]

~~40. NAME: Mr. Magne Albrigtsen COUNTRY: Norway INSTITUTION: UNEP P.O. Box 47074, Nairobi TEL: 230800 Ext 6004 FAX: 226890~~

41. NAME: Mrs. 0. V. Mejabi COUNTRY: Nigeria INSTITUTION: University of Ilorin Computer Centre, University of Ilorin, P.M.B. 1515, Ilorin TEL: 031 221090 FAX: 031 223511 E-mail: relay through April Jones of McMaster University, Canada on [email protected]

42. NAME: Dr. S. Ramani COUNTRY: India INSTITUTION: National Centre for Software Technology NCST, Gulmohar Cross Road Nog Juhu, Bombay 400 049, India TEL: + 91 22 620 0590 FAX: + 91 22 621 0139 E-mail [email protected]

43. NAME: Mr. William H. Sangiwa COUNTRY: Tanzania INSTITUTION: Muhimbili Medical Centre P.O. Box 65015 Dar es Salaam, Tanzania TEL: + 255 51 27081/6 ext 240 FAX: +255-51 46163 E-mail 5:733/1 (fidonet) [email protected] (INTERNET)

~~196 44. NAME: Dr. Douglas Rigby COUNTRY: Kenya INSTITUTION: ELCI/Satellife P.O. Box 72461 Nairobi TEL: 254-2 562015/22 FAX: 254 2 562175 E-mail: [email protected] 5:731/1~~

45. NAME: Mr. John Schoneboom COUNTRY: USA of Science INSTITUTION: American Association for the Advancement 1333 H Street NW, Washington, DC 20005 TEL: 202/326 6651 FAX: 202/289 4958 E-mail- [email protected]

~~46. NAME: Mr. Rushby COUNTRY: Pakistan INSTITUTION: UNHCR c/o UNHCR, case Postal 2300 Geneva Drpot 2 Switzerland TEL: 92 51 212051 FAX: 92 51 212058~~

~~47. NAME: Mr. Tchepgag Justin COUNTRY: Cameroon INSTITUTION: Ministry of Posts and Telecoms CETCAM - DIRTELCAM Ministry of P&T Yaounde, Republic of Cameroon TEL: 237 23 57 91 FAX: 237 23 16 63~~

~~48. NAME: Mr. Dewayne Hendricks COUNTRY: USA INSTITUTION: Tetherless Access Ltd 43730 VISTA DEL MAR, Fremont, CA 94539 TEL: 510 659-0809 FAX: 510 770-9854 E-mail: [email protected]~~

~~49. NAME: Dr. C.J. Buffam COUNTRY: Switzerland INSTITUTION: U.N.H.C.R. 5-7 Av De La Paix, CH 1202, Geneva TEL: +41 22 739 8185 FAX: +41 22 731 95 46~~

197 50. NAME: Mr. William S. Shu COUNTRY: Cameroon INSTITUTION: University of Yaounde Dpt of Computer Science, University of Yaounde, BP 812 Yaounde, Cameroon TEL: + 237 23 1602 Ext 1159 FAX: + 237 23 5588 /+ 237 22 1320

~~51. NAME: Prof. P.E. Mugambi COUNTRY: Uganda INSTITUTION: Makerere University P.O. Box 7062. Kampala, Uganda TEL: 532401/532440/559712 FAX: 041 533440 E-mail: Paul. [email protected]~~

~~52. NAME: Mr. F.K. Mbua COUNTRY: Kenya INSTITUTION: KPTC P.O. Box 30304, Nairobi TEL: 229061~~

~~53. NAME: Mr. Francis Mwasi COUNTRY: Kenya INSTITUTION: KP&TC P.O. Box 30301, Nairobi TEL: +254-2227401 Ext 3279 FAX: +254-2336379~~

~~55. NAME: Prof. T.S. Dod COUNTRY: Kenya INSTITUTION: Kenyatta University P.O. Box 43844, Nairobi TEL: 810901 Ext 340 FAX: 810759~~

~~56. NAME: Dr. Kewamoi Sogomo COUNTRY: Kenya INSTITUTION: Kenyatta University P.O. Box 43844, Nairobi TEL: 810901 Ext 326 FAX: 810759~~

~~57. NAME: Dr. G.M. Macharia COUNTRY: Kenya INSTITUTION Institute of Computer Science University of Nairobi P.O. Box 30197, Nairobi TEL: 442121 Ext 335 FAX 447870 E-mail [email protected]~~

~~193 58. NAME: Dr. S.C. Chhabra COUNTRY: Kenya INSTITUTION: Kenyatta University P.O. Box 43844, Nairobi TEL: 810901 Ext 355 FAX: 810759~~

~~59. NAME: Dr. Shem Ochuodho COUNTRY: Kenya INSTITUTION: University of Nairobi P.O. Box 30197, Nairobi TEL: 442121 Ext 335 FAX: E-mail [email protected]~~

~~60. NAME: Mr. Micheal Hailu COUNTRY: Kenya INSTITUTION: ICRAF P.O. Box 30677, Nairobi, Kenya TEL: 254-2-521450 FAX: 521001 E-mail: [email protected]~~

~~61. NAME: Dr. Kinuthia R. Ngugl COUNTRY: Kenya INSTITUTION: University of Nairobi P.O. Box 34336, Nairobi TEL: 632824/5~~

~~62. NAME: Mike Jensen COUNTRY: South Africa INSTITUTION: Greennet/Worknet/Ngonet P.O. Box 18866, Hillbrow 2038, South Africa TEL: 27-11-484-3777 FAX: 27-11-484-3557 E-mail [email protected]~~

~~63. NAME: Mr. Phil Gray COUNTRY: Mozambique INSTITUTION: Care International 660 first ave., New York NY10016 TEL: (Mozambique) 258-1-492064/5/6 FAX: 258-1-492077~~

~~64. NAME: Harun N. Baiya COUNTRY: Kenya INSTITUTION: Volunteers In Technical Assistance (VITA) P.O. Box 34336, Nairobi TEL/FAX: 721872~~

~~199 200 Additional Papers~~

~~201 202 United Nations University~~

~~The United Nations University~~

~~Introduction~~

The United Nations University (UNU) was proposed by the UN Secretary General U Thant in 1969. After world-wide consultations it was recommended that a new kind of university should be established, designed to:

promote international scholarly and scientific co-operation in helping to solve urgent global concerns of survival and welfare b y undertaking problem-oriented, multidisciplinary research through world-wide networks of institutions and individuals and by strengthening research (and advanced training) capabilities in developing countries.

~~The UNU is organized on the network principle, which by itself calls for efficient communications within the network, having the following component parts:~~

~~University Centre in Tokyo - central programming, co- ordinating and administrative body~~

~~Research and Training Centres - created by the UNU itself in various parts of. the world~~

~~Associations with existing universities, and national (regional) research centres, located mainly but not exclusively, in developing countries.~~

Together, these elements interact to create a flexible and organic University system. The opening words of the UNU Charter (adopted by the General assembly of the United Nations on 6 December 1973) define the UNU as an international community of scholars. 203 United Nations University

Through its world-wide activities, the UNU has "its location at the site of each centre or programme". Hence, it is clear that the "life- blood" of the UNU are its communications: the functioning of the UNU system is by definition dependent on communicating and disseminating (knowledge and experiences). This has been clearly stated in the UNU Charter where "dissemination" of knowledge is designated as an activity of the same order of importance as research and advanced training.

As mentioned, in addition to the networks of participating institutions (universities and research centres) and individual scholars, the UNU establishes its own institutes to deal with long-term problems that are not sufficiently being addressed elsewhere.

The first such UNU Research and Training Centre (RTC) is the World Institute for Development Economics Research (WIDER) in Helsinki, Finland, (1985) whose principal purpose is to deal with policy-oriented socio-economic research on development problems.

The secondly established is the Institute for New Technologies (I N T E C H) in Maastricht, Netherlands, (1990) where studies are undertaken on the impact of the modern technologies and their socioeconomic implications, especially in the developing countries (1).

The newest centre is the International Institute for Software Technology (IIST), in Macau, South East Asia, (1992). Its purpose is to undertake research and training on the development and adaptation of advanced computer software , especially to strengthen the capabilities of developing countries in software technology and the management of software projects, and provide them with opportunities to participate internationally in current high quality, relevant research and development.

In line with the other UNU centres and programmes, the UNU/IIST intends to develop an 'Organic Network" consisting of an expanding circle of affiliated or co-operating software centres and research institutes, which will focus on the developing countries. The UNU/IIST will assist other members of its network in becoming connected to an electronic 'e-mail" service which will form the backbone of the network. Uniform electronic network access will be provided to each affiliated institution so as to avoid a division into "first" class and "second" class affiliates, be they in academia or industry. Obviously, the IIST considers communicating in research and disseminating knowledge, as essential in order to fulfill its overall goals and objectives ( 2). 204 United Nations University

Other UNU RTC's are now at different planning stages and will be in the areas of: governance, natural resources, biotechnology and marine sciences. An Institute of Advanced Studies in Japan will be established next to the UNU headquarters building in the Tokyo metropolitan area.

~~Perhaps the questions asked most often about the UNU are:~~

~~Where is the caml2us. who are the students and the faculty members ? How to apply to enter this new University?~~

There are no students in the usual sense when thinking of enrolled undergraduates or regular post-graduates working towards academic degrees. The UNU "fellows" are young scientists and professionals, mainly from the developing countries who are engaged in UNU research and advanced training activities, this training being "training for research" and "training through research" i.e. problem or task oriented. The training period is of varied duration, anything from a few months to up to two years at maximum.

Strictly speaking, a potential fellow cannot apply for admission to training. The UNU fellows are chosen by personal interviews after recommendation from their home institution, which must be working as partner of the UNU or in an area of priority concern to the UNU Programme. Candidates must be committed to returning to work at their home institutions, which must guarantee them an opportunity to use their newly acquired knowledge and skills on their return. Usually, the UNU organizes a follow-up of its past fellows endeavoring to remain in contact with them and their home institutions. Here again, efficient easily accessible means for ensuring continuous communications are needed.

Thus, the UNU scholarly community consists of all those - researchers, scientists, engineers and university teachers, who are engaged in this global partnership of research and training anywhere in the world. They must be linked together in a flexible and efficient UNU communication network, consisting of state-of-the-art sophisticated as well as of low-cost simple networking segments (sub-systems).

Financial suQl2ort for the UNU comes entirely from voluntary contributions from governments, bilateral and multilateral agencies, foundations, and other public and private sources. UNU does not receive 205 United Nations University any funds from the regular budget of the United Nations. Major contributions are to its endowment fund, a capital fund invested to yield the basic annual income i. e. the core budget. The UNU also receives annual-type operating contributions as well as specific programme or project financial support.

~~In the 1990s the UNU programme activities cover five major areas of global concern:~~

~~Peace, Governance and Culture~~

~~The Global Economy~~

~~Global Life-Support Systems~~

~~Population, Health and Human Welfare~~

Science and Technology, a research area which comprises the consequences of technological changes at the frontiers of knowledge - informatics, microelectronics, molecular biology, new materials - which are reweaving the fabric of the contemporary global society and will shape the future of mankind in the next century . In addition, the University carries out activities aimed at e x p a n d i n g access world-wide to scientific information and promote use of new information/communication technologies (3).

~~In this area is situated the UNU: Microprocessors and Informatics Programme~~

~~BACKGROUND~~

This programme was launched by the United Nations University (UNU) in 1983 under the heading of "Mastering of Microprocessor Technology". Its main goal was to respond to the concerns repeatedly expressed by the developing countries, facing the challenge of the new information technology (informatics), which is powerful, all-pervasive and producing far-flung global impacts. The world is aware of the fact that industrialized countries control this essentially "science- based" technology. At the same time, informatics is rapidly changing world economies, and the social life of communities world-wide, while reinforcing the traditional striking advantages in science and technology applications of the advanced countries over the less developed ones.

~~206 United Nations University~~

The UNU considered that as being a global problem : it is crucial for global development that all the nations and countries should "master" this new technology and its applications. Indeed, it is inequitable that only one restricted part of mankind should be in full possession of information technology . Even more so, it is inequitable and unproductive that to one large part of mankind should be denied the right to face this challenge and to actively contribute to the advancement of the complex of "informatics". In accordance with its mandate, the UNU reacted to this global challenge.

The lack of capacity to contribute to the progress of informatics is obviously caused by many factors, but the UNU considered that the deficiencies of the universities in the developing countries were one of the crucial obstacles to world- wide progress and implementation of informatics.

It became clear from the start that the first task of the UNU was to aim at strengthening the scientific and academic communities in developing countries to enable them to help their societies to make better use of informatics for development. It was expected that by concentrating on carefully selected locations and sponsoring development-oriented research and training, the UNU would help build active and competent groups of specialists in developing countries, who would play the role of initiators and innovators. The UNU would provide means for expanding the influence of these specialists over the local levels so as to make them active contributors to regional and global development through informatics, al pari with their colleagues from the industrialized countries.

~~OBJECTIVES~~

Therefore the main objective of the UNU Informatics programme is to help universities world-wide, particularly in developing countries, to strengthen their scientific (research and teaching) and technical (innovation) capabilities in this field primarily by:

~~A. assisting them to initiate and improve relevant research and training in their home institutions, and~~

~~B. enabling them to collaborate world-wide with colleagues both in other developing countries and in the industrialized countries.~~

~~Systematic action is undertaken to support the universities to:~~

~~t.- set up competent local teams of researchers linked to international scientific networks consisting of front-end research centres, and~~

~~2.- organize advanced teaching (with visiting professors) awarding academic degrees generally recognized as equal to those obtained at the universities in the industrially developed countries.~~

At the same time, it is crucial to enable universities to reach out to the various partners involved in the development process of the society, in order IQ make the universities more responsive to their countries overall develoRm ental

~~207 United Nations University needs and efforts. Only in this case will the universities enjoy full support of their local authorities and the general public.~~

In this respect, it is essential to ensure that the universities acquire all the necessary communication means for reaching out to other partners in their respective countries, but also to get permanent links with relevant groups of researchers and teachers abroad. Thus, development of communications is an integral part of the UNU informatics Rrogramme , at the planning stage and during the conduct of work and the final evaluation.

UNU sponsored research, "training for research" and "training of trainers", as well as dissemination of information, are continuing along these lines. They endeavor to respond to the increasing requirements of universities in the developing countries, which are eager to upgrade the capabilities of their academic and technical personnel and to provide advanced teaching to their students and postgraduates. This is considered, by the universities and their constituencies in the developing countries to be the prerequisite of "self-reliance" in informatics which is one basic condition for over-all social development in the coming century. Needless to say that the UNU fully shares these views, but does not impose them from outside.

~~ACTIVITIES~~

~~In 1990-1995 the UNU Informatics Programme is focussing on three areas:~~

~~design of real-time control systems, research and advanced teaching of informatics, Informatics for system design and management,~~

~~most activities coordinated by the UNU being carried out jointly with other universities and research institutions throughout the world, especially in the developing countries .~~

~~1. - design of real-time control systems, Le. integrated microprocessor- controlled Instrumentation:~~

Research and training at the Microprocessor Laboratory in Trieste, in co-operation with the International Centre for Theoretical Physics (ICTP) IAEA, is based on a grant to UNU from the Government of Italy. The Microprocessor Laboratory , as part of an institute whose main interest is scientific, selects to solve technological problems embedded within problems which have scientific relevance. It tries to avoid choosing problems that fit technologies already mastered elsewhere. By all means , it tries to avoid becoming an electronic shop or service where developments are carried out "on request". In every case, only after the theoretical aspects have been solved, a particular technology or architecture is chosen. The Laboratory has cutting-edge technological alternatives for the design of instrumentation: from classical board-level design using microprocessors or programmable hardware logic, to the design of integrated circuits, be they gate 208 United Nations University arrays, optimized gate arrays or full custom design. During the whole process of R&D, the benefits of computer design tools are emphasized, enhancing the group problem solving ability of the participating scientists, mostly from -the developing countries.

International colleges (in Trieste) and regional training courses in Latin America, Asia and Africa are organized in cooperation with several universities in Italy and other European countries as well as with the CERN , Geneva. Assistance is provided to scientific instrumentation laboratories and microprocessor-support units at universities in Latin America and Africa.

This part of the UNU programme on informatics, linking UNU and IAEA, has been successfully carried on since 1983. Over one thousand of scientists and engineers from the developing countries have attended various colleges and training courses reaching from basics of microprocessors technology and applications in scientific instrumentation to advanced teaching in VLSI Design Techniques, the Design of Real-Time Control Systems and in Telematics.

Research teams from developing countries have actively contributed to investigations in these advanced fields of informatics. Projects are carried out using cutting-edge technology since competition in high technology is global. No concession is made in this respect to "appropriate technologies" since microprocessors, present the very special characteristic that the most advanced implementation of a solution is generally the most economical, compact, reliable and appropriate. Hence, only up-to-date solutions are admitted in this field. In the limits of the possible, the Laboratory tries to use hardware and software readily to be provided to the developing countries. Computer networking through the world- wide electronic mail and file-sharing systems is fully realized , connecting also many countries in Latin America and Asia, though still less in Africa.

~~2. - research and advanced teaching of informatics (computer science) in selected universities combined with out-reach activities:~~

In cooperation with the "Institut National de Recherche en Informatique et en Automatique (INRIA)", Paris, advanced academic teaching in informatics and front-end research on parallelism, data-base management systems and networking, are proceeding at the University of Yaounde in Cameroon. This joined project has started in 1985. Presently it is extended to other universities in Western and Central Africa, thus contributing to the establishment of a regional research and training network in informatics. In fall 1992, the "First African Conference on Research in Computer Science' will be organized with the support of the Government of France. Participation will be exchanged with the XVIII Latin- American Informatics Conference PANEL'92. In the following period common research projects and secure communication links will be established between African, Latin American and Asian universities, with the support of the scientific institutes and the research funds in France and Spain.

In cooperation with universities in Ireland, modalities are being developed in several African countries ( Ethiopia, Sudan, Tanzania, Uganda, Zimbabwe, Zambia ) for linking academic research with actual system design, management and production control in local industries and in services. In order 209 United Nations University to carry out their work, the specialists obviously require continuous up-dating of knowledge and skills in this rapidly expanding area of science and technology.

The lack of a critical mass in most developing countries and the general constraints to circulation of scientists, constitute severe obstacles for technological progress and innovation in this high-tech field which has great developmental potentials. For this purpose specialized training , including study periods in the technology parks of Irish universities, is provided to a number of researchers from Africa. Regional workshops in Africa will be organized in 1992 and 1993 to propagate the ideas, explore the needs, exchange the experiences and support dissemination of research results. They will also serve to initiate co-operative work by helping to identify prospective candidates for future teams of local/regional R&D, especially those requiring further training in Ireland.

The dissemination of knowledge through provision of stable information flows, using modern communication means including electronic mail when appropriate, and the circulation/publication of research results, are foreseen. This will include also such events as the mentioned regional workshops in Africa .

This part of the UNU informatics programme is the extension of a successful project of cooperation between the UNU and Trinity College Dublin (TCD), which started in 1985 supported by a grant of the Government of Ireland.

~~3. - informatics for system design and management of complex man-made systems:~~

The Information and Decision Systems (INDES) project promotes research in information science and technology, towards a set of specific constraints prevailing in developing countries, which prevent them to use existing advances in information technology for the design and management of complex man-made systems; more particularly to integrate their systems of vital importance for development, with the international (global) systems. Plans are under way to initiate major activities in the area of information processing for decision support in complex large scale systems e.g. transportation, tourism, water management, food provision, health care and welfare.

In order to devise a scheduled implementation of the UNU INDIES project, it has been necessary to first convene in 1989 a Round Table of experts for clarifying the technical issues which are crucial for the success of any research undertaken in the field of informatics and information systems, research necessarily restricted to what seem to be attainable goals.

~~210 United Nations University~~

The mentioned INDES Round Table reached the consensus that the complexity of decision-making in general, and especially when applied to large scale man-made systems, is such that a comprehensive approach to its solution is first not available. Thus , it was considered that the preferred approach is to focus on the information processing aspects, since the availability of adequate information is a necessary precondition for effective decision making. Subsequently, a Workshop organized in 1990 on "INDES Phase I: Information Processing for Decision Making" produced a set of research directions in response to the charge given to the participants, namely to identify relevant information technologies and formulate a research programme on information processing for decision making in complex man-made systems: which is the fir focus of the UNU INDES project.

Thus, in addition to what is being implemented in all the UNU programmes, namely the integration of information and communication activities ( "dissemination of knowledge" according to the UNU Charter) with research and advanced training, in the INDES project information itself i.e. information processing , information transfer, information use, is treated as the subject of study. Obviously, "computers and communications" are at the heart of the problem.

Recently, in May 1992, the UNU has organized with the University of Kyoto, Japan, the Second International Symposium in the series "Frontiers of Science and Technology" entitled EXPANDING ACCESS TO SCIENCE AND TECHNOLOGY - THE ROLE OF INFORMATION TECHNOLOGIES.

Its purpose was to assess the potential of new technologies in providing access to, retrieval, handling and exchange of information. Intelligent access to information and its impact on communication as well as opportunities of developing countries in utilizing the new information technologies have been of major concern.

At the Symposium, the importance of international cooperation has been underlined. New modalities of co-operation, based on the potential of the current and future technologies have been suggested. Follow-up activities are now being proposed by the UNU to its partners, especially in view of enhancing international co-operation to ensure world-wide the equity of access to science and technology. A more close collaboration of all the international agencies and the relevant national bodies, public and private, is considered as crucial for success in what should be a global endeavor.

~~211 United Nations University~~

~~CONCLUSION~~

~~There are at least three good reasons why the UNU should be involved in communications study and development as a -Priority~~

1. The UNU is a "world-wide system", a network of research and postgraduate training centres and programmes, hence communicating within this system as a precondition for its existence and functioning;

2. The UNU's mandate is to "engage in research, post-graduate training and dissemination of knowledge", hence communicatina has to be considered as a task of the same order of importance as research and training;

3. The UNU has as "a central objective the continuing growth of vigorous academic and scientific communities everywhere and particularly in the developing countries, devoted to their vital needs .. " this is a global issue, hence study of the mechanisms of information handling and transfer i.e. communicating., is itself a part of the UNU's research programme as exemplified in its projects dealing with Informatics.

~~So far, the UNU has fully adhered to these principles and will continue in the future.~~

~~REFERENCES~~

~~1. United Nations University, INTECH,Institute for New Technologies, Tokyo, 1991~~

~~2. Bjorner, D., A Role for UNUMIST: Developing Countries' Access to New Information Technologies, UNU Symposium Kyoto'92~~

~~3. United Nations University, Research and Training, Tokyo, 1991~~

~~212 AFRICAN ACADEMY OF SCIENCES/AMERICAN ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE WORKSHOP ON~~

~~SCIENCE AND TECHNOLOGY COMMUNICATION NETWORKS IN AFRICA 27-29 AUGUST, 1992,~~

~~SAFARI PARK HOTEL, NAIROBI, KENYA~~

~~THE BIG OPPORTUNITY: THE CASE OF HEALTHNET IN ZAMBIA~~

~~REGINA CAMMY SHAKAKATA~~

~~MEDICAL LIBRARIAN/ SATELLIFE SECRETARIAT FOR LIBRARIES IN AFRICA~~

~~UNIVERSITY OF ZAMBIA MEDICAL LIBRARY P.O. BOX 50110 10101 LUSAKA ZAMBIA~~

~~Tel: 260-1-250801 FAX: 260-1-253952 Telex: ZA 44370 Email: [email protected]~~

~~213 BACKGROUND~~

~~The Republic of Zambia, a landlocked country, has an area of 752,600 kilometers, 90% of which is habitable.~~

~~The country shares borders with Malawi, Mozambique, Zim- babwe, Namibia, Angola, Zaire, Tanzania and Botswana.~~

~~Administratively, the coutnry is divided into nine (9) Provinces and 59 Districts. The provinces are:- Northern, Copperbelt, North-western, Luapula, Eastern, Southern, Lusaka, Western and Central.~~

Major towns and provincial centres of the country are well served by all weather roads, but access to interior parts of the country is difficult epecially during the rainy season. The country continues to enjoy a relatively good telephone infrastructure. Zambia continues to depend heavily on its railway system for its exports and imports of commodities. The main outlets being Dar-es-Salaam - Tanzania and Durban - South Africa.

The population of Zambia now stands at 8.09 million (Central Statistics Office 1990 Census). There has been an increase of 92% since 1969 census when the population was 4.06 million. The growth rate is 3.2% per annum for the period 1980-1990. The growth is largely due to natural increase. Population distribution varies considerably, with a national population density of 11.1 per sqaure kilometer, while the provincial densities range from 3.0 (North-Western to 55.2 (Lusaka) per square kilometer. Much of this variation is the result of urbanisation., with high densities generally occuring around urban centres. In rural areas, the most densely populated areas aie those along the borderlines of Luapula Province and Zaire, Eastern Province and Malawi and Northern Province and Tanzania.

The normal pattern of the rural habitation is in small settlements and hamlets. This set-up makes the provision of health services difficult. The rurul urban drift continues especially to Lusaka and the industrial towns of the Copperbelt. The urban population. (42% in 1966 ` is growing at 7% per annum. The rest of the population (58%) lives in the rural areas.

~~In spite of successes in some areas of development, the nation has not been able to achieve its target in health services provision mainly because of the ever increasing~~

214 demands for health services both in urban and rural areas and the lack of resources to expand the services. Health demands are significantly made when there are epidemics such as cholera and of late the high incidents of AIDS has placed greater demands for health services. The inability by the nation to develop an effective health information system have frustrated the efforts of the sparsely distributed health workers to cope with the health problems. As a consequent, the introduction of HealthNet in Zambia was seen as a BIG OPPORTUNITY to overcome some of the communication related health problems.

The economy is dependent on copper which in 1984 accounted for over 92.6% of the total export. There is still insufficient food production and Zambia imports several food stuffs to meet her staple and nutritional requirements. During 1991-1992 rainy season, the country suffered the worst drought ever which affected food production and this frustrated efforts to be self sufficient in food.

Zambia's economy has been drastically affected by the falling Copper revenue. In addition to this are various other factors such as the country's participation in the libera- tion struggle in the region in which nearly K10 billion has ,been spent, land-locked position, world recession, high oil prices, the gulf crisis, drought and sometimes floods which have all contributed to this decline. The general cost of living shot up to its highest level during 1986 and still continues to rise following the decontrol of prices, scarci- ty of foreign exchange and reduction in subsidies to paras- tals by the Government and the continued devaluataion of the kwacha. For the last five (5) years, the frequent fluctua- tion of the rate of local currency has affected planning and budgeting. In October, 1985, the Government decided to introduce the auction system of foreign exchange in the country. As a result of this, the value of the kwacha to the US Dollar fell from the pre-auction rate of K2.20 to between K21.05 to 1 USD for the period October 1986 to April 1987. The kwacha kept falling and by end of 1990 the rate was K48 - K50 per 1 USD. By April 1992 the kwacha stood at K140.00 per 1 USD. Today, July, 1992 the value of the kwacha is K201 to 1 USD.

With this type of economic background one should understand the circumstances under which health workers and institutions exist in Zambia. The ravaged economy has not spared the health sector either. The distant medical training and the supply of literature to medical students have suffered severe disruptions due to poor and dear communication facil-

~~215 ities available in the country. The BIG OPPORTUNITY offered by HealthNet was received like manna from heaven by Zambia.~~

~~HEALTH INFORMATION DILEMMAS IN ZAMBIA One of the primary ways that defines "Developing Countries" (DCs) as having need for developing is their lack of access to information.~~

Given the supposition that 75-80% of all relevant information today has been developed within the last two decades, it is clearly true that only those with access to latest research and thinking are going to be able to take their place in the international community. In the DGs this is not the case.

It is believed, for example, that at least 90% of data on Africa is thought to reside on databases in the North. This, not only means that Africa is amongst the 'have-nots" in terms of wealth, but also the 'have-nots' in terms of information. Economic and political competition are biased against her because of the weight of knowledge (even about herself) that is against her, and to which she does not have access.

If this is true at Africa region level, it is equally true at Zambia national level, as many of the basic methods for dissemination of information in the North (mass media; libraries, databases; telecommunications, etc.) do not apply in the South.

Medical Libraries: It is possible to think of many examples in the medical sector. Medical Literature is amongst the most expensive there is, partly because of its rapidly changing nature. Medical libraries cannot therefore afford to stock-the volume of books they require. Journals are not only expensive, but require regualr access to foreign exchange in order to maintain subscriptions. In some cases the health institutions are small in size and cannot afford the range of literature required. Rapid devaluation of the local currency also means that budgeting for literature is close to impossible, and the expense is very high incomparison to local costs, such as labour and basic drugs.

~~In addition, the alternative to keeping large volumes of paper literature, for example requesting specific information from colleagues in related fields has severe practical~~

~~216 drawbacks. convetional mail is very slow and unreliable. fax More modern electronic means of communication such as may be prohibitively expensive.~~

Ministry of Health: for the Ministry of Health generates numerous bio-data The This districts, provinces and the national headquarters. manually, in more than 90% ofthe cases is compiled bio-data from a serious question on the statistics emanating causing the these sources. The other data that is generated by Health is in the areas of planning, budgeting, Ministry of HQ accounting and monitoring data between Planning Unit (PU) Other types of and District Health Mnagement Team (DHM). data generated by the Ministry of Health are administrative reports, data, drug requisition,stock records, monthly Unit (HIU), epidemiological data between Health Information the health statistics returns and reports etc. Although the telephone service ranks better than most countries in region, still the high cost of ringing and the difficiency retard of and lack of provision of substantive information transmission of international information. Conse- speedy to quently rendering the Ministeries of Helth less efficient manage health affairs.

The World Health Organisation: World Health Organisation (WHO) generated information on the health other hand is voluminous and its dissemination to all communi- units is slowed by lack of an inexpensive reliable cation service. The existing systems including surface to travel are both prohibitively expensive and tideous domi- operate. While it is not intentional that WHO would is nate the use of HealthNet, its participating in HealthNet health creditable because it is the single generator of most information of both local and global interest.

Medical Training: Medical training is both formal. through schools of medicine and colleges, and informal through supervised distance learning and instruction on projects, assignments, background data, submission of student reports etc. and comments from teachers and research supervisors. The promotion of availablity of journals, continuning education, specific postgraduate training programmes, undergradute training and patient care have been done at great cost energy-wise and economically.. Other programmes which are costly to manage currently are planning working visits by consultants and the enforcing of special drug programmes (policies) or investigations facilities.

217 Professional Journals: The publishing of professional journals like the MEDICAL JOURNAL OF ZAMBIA has beeen erratic with gaps of up to four years i.e. 1986-1990. The only mouthpiece the health the medical professionals have to present _their views has suffered serious financial setbacks. What is required is a system to alleviate the publishing problems that have be- sieged the journal in the past i.e. publishing elctronical- ly.

The introduction of eletronic mail (and associated conferencing and literature distribution can clearly go some way towrds alleviating the above problems. In Zambia,whilst the development of email and in particular HealthNet is at very embryonic stage measured across the continent as a whole, it certainly appears to be a potentially important vehicle for the dissemination and exchange of information both on a south-south and South-North basis. Rapid development ofthis technology is likely in Zambia because of the good telecommunication infrastructure. The introduction of satellite technology, and more robust telecommunication apparatus will shortly make email more widely avaiable particularly to the health, non-governmental and academic communities.

HEALTHNET INFORMATION SERVICES SatelLife administers HealthNet, a telecommunication system designed to facilitate the exchange of information among health professionals in developing world countries and to link them to their collegues in the North. HealthNet has two problems which it anticipates to solve in the developing world-:

~~1. the problems which the physcians, health scientists and health-care workers encourter when attemping to communicate with each other and~~

2. the isolation these people experience as a result of their lack of access to current sources of information about health problems in their countries and about global health issues which affect them. Zambia, unlike most of Africa has good telecommunication systems although when used internationally they tend to be very expensive. There is need in Zambia and the rest of Africa to have electronic mail which is not dependent on traditional communications infrastructures. For example, in Zambia, an international call costs USD 6.00 per minute. In Kenya, a fax costs USD 7.70 per page outgoing. In Tanzania, the minimal cost of a telex costs more than USD 25.00.

~~218 These prohibitive communication charges lead to the isolation of students; researchers, physicians and healthcare professionals from each other and from their colleagues around the world.~~

As a consequent, the objective of HealthNet is to provide physicians, health scientists and healthcare workers -in Africa with a reliable, inexpensive telecommunication system. This service gives them access to each other and their collegues in the developed world, to electronic mail networks outside Africa, to remote databases and to affordable and relevant medical literature.

~~The HealthNet Information Services provided by SatelLife include:-~~

1. Library Partnership Programmes: In this programme, medical libraries in developing world countries and those in industrial countries create partnerships for the purpose of requesting and exchanging health information. The Universi- ty of Zambia Medical Library has twinned with the University of Florida Medical Library. The two libraries are already exchanging health information, and the system is proving beneficial, especially to the University of Zambia Medical Library.

2. HealthNet News: This electronic publication is transmitted over HealthNet and contains full text articles, journal articles summaries, and commissioned articles. Although, at the moment, the HEALTHNET NEWS is only dissemi- nated to health workers within the School of Medicine and the University Teaching Hospital, it is planned to circulate it to all the health institutions once the directory of health institutions is compiled.

3. Consultation: HeathNet provides physicians and healthcare workers in developing world countries with on-call access to organizations and institutions. Initial participating organizations include the Harvard School of Public Health, the Liverpool School of Tropical Medicine, and the Oswaldo Cruz Institute in Brazil. In Africa, consultations have taken place among the Universities of Zambia, Dar es Salaam, Mozambique and. Makerere. This is encouraging, however, there is need to stimulate interest among health wokers to use the HealthNet system more than is currently the case. Intranationally, the institutions that are talking to each other are the Tropical Diseases Research Centre (T D R C ), W H 0, Medical Library, School of Medicine the Ministry of Health and the Zambia Consolidated Copper Mines (Z C CM).

219 4. Research Capabilities: CD-ROM drives are being installed in conjunction with ground stations to allow initial research on MEDLINE and archival retrieval of abstracts to be performed on site. A CD-ROM station has just been installed at the University of Zambia Medical Library and al ready it has stimulated alot of interest among library users who are making numerous literature search requests. Through HealthNet and access to MEDLINE on CD-ROM, it is anticipated to generate a publication called the DIGEST. Consequently, a DIRECTORY OF HEALTH PROFESSIONALS IN ZAMBIA will be compiled. These professionals will be on the DIGEST and other health mailing lists.

These services are and will be made possible through the use of both a "store and forward" satellite called HealthSat and a terrestrial system which is supported by the P T C that pick up and deliver eletronic mail messages. Because HealthSat does not depend on international telecommunication links, it is not affected by congested circuits, unaffordable service charges or frequently disrupted services.

THE STATUS OF HEALTHNET IN ZAMBIA TODAY In a world in which 90% of data on Zambia is thought to reside on databases in the North, information poverty clearly contributes to the economic and political disadvantges that Zambia faces. When economies run into trouble the effects on basic health care are often severe, and are compounded by a dear communications infrastructure and the lack of the very information that the medical sector needs to tackle the problems it faces. Medical journals and books are beyond the budgets of most institutions even if the foreign exchange that they require was available.

Internationally HealthNet has been developed to address this issue by giving countries in the 'South' access to satellite facilities which link them to medical information reposito- ries (libraries, health institutions and universities) in the 'North' as well to their with whom thety share common problems and goa s. In zam ia, at interna ions level telephone systems tend to be good but the costs are set artificially high in order to dissuade outgoing calls which incur international carrier charges in foreign exchange. Thus, while international email links have been established between the University of Zambia and the rest of the world, Healthnet provides a cheaper and more reliable alternative satellite technlogy.

220 Apart from the existing connections on the HealthNet system, there is a proposal for a pilot project to be carried out within just one province (Southern Province), its aim being to install electronic mail in seven out of the eleven district hospitals, and then over a six month period to monitor closely the performance and uses of email in those hospitals. The seven target hospitals are:- Choma General Hospital, Macha Mission Hospital(Choma District), Living- stone General Hospital, Chikankata Hospital(Mazabuka Dis- trict Hospital, Monze District Hospital and Siavonga Dis- trict Hospital.

A single province was slected to promote as much realisitc interaction between the paticipants as possible. However, it is important that links with regualr contacts in Lusaka should also be in place. While the Ministry of Health, the School of Medicine at the University Teaching Hospital (UTH), WHO Lusaka and UNICEF all have email, part of the project will involve the identification of other of these Lusaka-based contacts.

While district hospitals outside of Southern Province who already have the required equipment (Computer,modem and adjacent telephone line), in particular those involved in the New National Health Policy Pilot Project, will also be encourgaed to participate in HealthNet, during this project these districts will not necessarily receive same level of support to the seven target hospitals.

The project requirements can be broken down into four main areas: the determination of materials requirements; an assessment of the technical. support needed; an analysis of the user skills involved, and the monitoring of usage. The details of the above requirements are contained. in the proposal for the expansion of HealthNet in Zambia.

~~USES OF HEALTHNET IN ZAMBIA~~

Many %-. have been identified within which the use of electronic mail through Heal thnet will be beneficial. The system has the potential to greatly improve access to people and information residing outside Zambia, as well as improving national communication on administrative professional and education matters. No doubt other ereas will be discovered once the system is being fully utilised.

~~At intranatioal level the links provided through the network's host computer both those of a terrestrial nature and~~

221 the satellite facility, will give health workers in the field access, to* the same contacts, information and data that HealthNet points in Lusaka and the Copperbelt currently enjoyed. Thus they will be able to access the wealth of health literature stored in health science libraries and information centents in Europe and American,thereby improving both motivation and medical competence. They will also be able to access information stored electronically in 'local' libraries e.g the Medical Library at the University

Teaching Hospital. In additional they will be able to receive accurate, up-to-date and relevant epidemiological data for Zambia. These HealthNet advantages will be supplemented by a greater ability to communication with fellow health workers in adjacent countries i.e.Mozambique and Tanzania, using the satellite link, thus enabling better clinical consultant and cooperation within Southern Africa.

Under the new National Health Policy improved communication between the Ministry of Health Headquarters and peripheral health service and management units and between the semi- autonomous peripheral units themselves will be very important. Much of this communciation will be administrative involving the exchange of planning, budgeting, accounting and manpower data; the provision of regular health statistics, returns and reports, and the redistribution of these reports; and the ordering and co-ordination of drugs, vehicle parts and other supplies.

Electronic mail has the potential to significantly improve communication on professional matters. Improved clinical consultation will be possible between district hospitals, and between centrally based specilists and peripheral doctors enabling greatly enhanced patient care especially in those district hospitals previously remote and 'inaccessible.' There will be more efficient disbursal of information regarding treatment facility and clinical management policy changes while advance warning of referrals and consultation visits will enable better management and preparation within the receiving hospital. It will also be possible to gather tldteo then equally quickly olais ribaute a linltgeorestle parties

For training, the use of email for distance and maintaining contact with mastesr students working as District Medical Officers will prove invaluable to the School of Medicine at UTH. In addition it will be possible to provide expanded education directed at all cadres of the health work including undergraduate students working in the districts. Such

~~222. continuing education is currently inadequate and largely concerned with medical graduates.~~

For improvement of library services, the medical Library at the UTH has twinned with the University of Florida Medical Library and the two libraries have been exchanging information. With the availability of MEDLINE on CD-ROM, the Medial Library at UTH will be able to conduct literature searches for the School of Medicine and for other health workers in the country. The Medical Library will also assist other health sciences libraries in the country by providing them with the DIGEST which can be used to scan for literature on popular medical subjects. Through the Medical Library, these libraries will enjoy the same facilities enjoyed by the Medical Library at UTH.

PARTICIPATION IN HEALTHNET In zambia, users of HealthNet have been categorised as medical libraries, the Ministry of Health, School of Nurs- ing, medical schools, WHO and other health care related institutions. These include the Churches Medical Associa- tion of Zambia (CMAZ), the Zambia Consolitated Copper Mines (ZCCM), the Zambia Physician for the Prevention of Nuclear War (ZPPNW), the Zambia Flying Doctor Service and medical research institutes i.e. the Tropical Diseases Research Centre.

Realising that there are several Nongovernmental Organisar- ion (NGOs) that carry out health programmes such as Kara House Counselling, Anti-AIDS Project, Zambia Association for Research and Development, Society of Women and AIDS in Zambia to name a few; that would benefit from HealthNet services; and aware of the difficulty of drawing a line on who should participate in HealthNet among the NGOs, an email network for NGOs (ZANGONET) was established. ZANGONET has a gateway to HealthNet. Therefore, interested NGOs would have access tc HealthNet through ZANGONET. HealthNet has gateways to international electronic mail networks such as BITNET. EARN, JANET, GREENET, ACADEMNET and NGONET. This feature allows HealthNet users in Zambia to send messges to institutions on other networks.

USERS EXPECTATIONS OF HEALTHNET The benefits.of the HealthNet information services are great especially to a country like Zambia that has searched for every opportunity to salvage its ailing health information services. The user expectation of HealthNet Information Services in:-

223 1. Library Partnership Programme are that in additional to medical libraries in developing world countries having to create partnerships for the purpose of requesting and exchanging health information, users would like to see develop a network of medical libraries users of HealthNet.

2. HealthNet News: Currently, the electronic publication container medical journal articles summaries and commissioned articles. Users would like to see the medical li-, brary component of the publication for purposes of providir.g current awareness on developments on information technology in medical library services.

3. Consultation: where as the international consultations among health workers is shaping up, there is need to strengthen similar consultations at intranational levels as well. This expectation implies greater supply of communication equipment such as personal computers to more health workers. This suggests greater financial implications for funding agencies.

4. Research Capabilities: Users of this facility expect to have CD-ROM drives to be installed in conjunction with ground stations to allow literature searching on MEDLINE and archival retrieval of abstracts to be performed on site. The extended expection of this service is that full articles would be sent by the same speedy system. There has been experiences whereby abstracts have been received but the full articles which were sent by ordinary mail did not arrive (10) weeks after despatch. This could be frustrating to HealthNet Users. The University to Zambia Medical Li- brary has had a CD-ROM Station installed on 17th July, 1992. Therefore, the need for full articles to be supplied becomes greater because the readers would have chosen what they want already.

5. The users of HealthNet expect to have a good supply of journals as a compliment to both the HealthNet and the CD- ROM MEDLINE searches. Unfortunately, this is not the case with the University cf Zambia Medical Library where the journal subscriptions have suffered as a consequent of the ailing national economy. However, there are some funding agencies that are willing to come to the aid of the Library.

Other user expectations would be known after the system has been widely used in the country. For now, HealthNet appears to be the answer to all health information related problems. Some policy makers would like to see a similar system develop to handle information other than health.

224 MANAGEMENT OF HEALTHNET IN ZAMBIA At global level,- HelathNet is manged by SatelLife. In Zambia HealthNet is managed by a Users Council which was set up by the Zambians themselves following advise from Satel- Life. The Council is composed of a cross section of the user communities. The composition includes the Medical Librarian, representatives of the ZPPNW, the University of Zambia Computer Centre and Medical School, the Ministry of Health, WHO, Medical Association of Zambia and Chainama Hills Hospital. An Ad-Hoc Planning Committee has been-set up to plan the expansion of HealthNet. It is chaired by-the Univeristy of Zambia School of Medicine. The Library is also serving as the HealthNet Liaison with Boston and as SatelLife Secretariat for Libraries in Africa.

~~225 BIBLIOGRAPHY~~

~~1. HealthNet News: v.1 (1) 1992.~~

~~2. Planned Parenthood Association of Zambia Country Situa- tion, 1992 (Unpublished)~~

~~3. Proposal for the expansion of HealthNet in Zambia, 1992 (Unpublished paper)~~

~~4. SatelLife. Parternship for global communication in Helath. Boston, [n.d.]~~

~~226 Comm~~

~~CEntral AMerican NETwork~~

~~William C. East PacComm Packet Radio Systems Tampa, Florida~~

The Central American Network (CEAMNET) net was originally worked out at the 1988 FRACAP meeting in San Jose Costa Rica.The idea was to link the 5 countries together in a packet network which would allow day to day amateur communications over the above network, and during emergencies the passage of emergency traffic between the areas affected and the rest, of the isthmus.After a bit of debate over what protocol to use, we settled on COSI, now known as ROSE or Rats Open System Environment.

~~The other networking systems studied were 1 (1) Digipeaters, (2) NetRom, (3) TexNet. our reasons for rejecting each one of the above systems is as follows:~~

Digipeaters, over one or two hops on channels with low usage these work quite well. But when you have very active channels and"hidden terminals" a digipeater network will rapidly come to a halt. Also the maximum path can be no more than 8 digipeaters in length.

NetRom, This was looked at as being a good answer to the problem. The cost of the EPROM's was a bit out of the reach of some parts of the network, but others were willing to make up the difference. The big problem for some countries was the lack of Exit point information available to any monitoring station, there was not a easy was to determine where the connecting stations were exiting (entering) the network. The countries who pointed out the illegality of NetRom were E1 Salvador and Guatemala.

TexNet was rejected due to the specialized controllers needed to run this networking system. It also suffered from the same end point identification that NetRom suffers from.We selected ROSE because it does identify the end points of a network link. It uses CCITT X.121 as the network addressing scheme, making it very easy to explain to the different PTT's involved in CEAM the network. We found before hand that 7uwt the aimgle PwC,o Poch* RM& 4413 X HwpuMn FL

~~227 explanation of AX-25 being a extension of X.25 made introduction of packet radio into many countries much easier.~~

For a good description of why central America selected ROSE as the regional networking system see the copy of my work published in the Sth ARRL Computer Networking Conference.The other reason for selecting ROSE was also based on the X.121 addressing. It makes it very easy to interface to the regional and national PAN's in the event of a emergency where a interlink may be needed. Our network in Costa Rica has grown to 5 switch sites which give coverage of approximately 808 of the country with minimal equipment. In many cases a hand held radio will get you into the nearest switch. In other cases the same handheld with a external antenna will get you to the switch. The 201 not covered is actually accessible from perhaps 958 of the country with a 10 watt radio and a antenna of some 7 - 10 db of gain-The other 5% can be covered with portable switches which the National Emergency Commission has available for just such a possible emergency.

We also installed and operate a Bulletin Board System (BBS) which carries information of general interest to the local amateur population, and also serves as a message system for exchanging messages between amateur stations, in the form of electronic mail. The BBS is also linked over HF frequencies to the international BBS network where much very interesting and important information moves back and forth between amateurs.

The proper selection of a networking system, along with the proper installation of the equipment allowed the Costa Rican Packet Network to survive, and cope with, a Richter 7.4 scale earthquake. 30 minutes after the major quake we were able to link into the network and check out it's integrity. All switches were on line, even a switch that was only about 40km from the epicenter. We were able to move into the affected area and set up portable packet stations and proceed to move data in and out of the area the next morning once the airport had been determined to be able to accept aircraft traffic.

All the network switches in Costa Rica are battery backed and many of them have battery power to last up to two weeks before needing external power. All sites are "seismic hardened", that is the building is either steel re-enforced concrete or steel plate construction. Since all sites do have commercial power, a secondary power source is not needed as sites will be powered up within hours of a failure, this being due to fact that all sites form a 228 e

~~important part of the national communications network. There are also broadcast stations, both radio and TV at these locations and the need to keep them on the air took on a great importance.~~

The system we used as a terminal program was a BBS program known as AA4RE BBS. This BBS program written by Roy Enguehausen, has a very powerful set of commands that allows the program to work very well in the amateur packet radio environment, yet avoid the problems of users not associated with the emergency at hand connecting to the BBS. Once a emergency hits been declared, the UBS system operator can close the BBS to all but emergency operations.

The systems installed in the disaster area was a BBS met up to operate as mail systems; all the operators had to know was how to type (within reason) and how to address mail (three addresses) So it was very easy to set the system up to take messages off of the keyboard and after the material had been formatted and the message closed, the system would then forward to the target system in the capital.

The network in other parts of Central America is being installed. Guatemala has 2 switches, Honduras 1 switch in service and is planning-to install another soon. E1 Salvador has one switch in service, Nicaragua has the equipment but needs to technical help with the set up. Panama has 2 switches of the wlhicb one is in service and the other is just awaiting some people to install the system.

The packet network has proved a real positive element in the day-to-day amateur operations; many amateurs who previously did not have good access to what was going on outside of their area, now can follow events in amateur and emergency activities on a day to day basis using the BBS. The same amateurs no longer have schedule meetings to on the radio, if their correspondent is not available, all the have to do is leave a message on packet for the distant station to read later when he/she is available.

Packet also allows emergency operations to make much better usage of radio frequencies, moving information and previously was too tedious to move or took too much time to move. The advent of the ROSE networking system can be passed on to the non-amateur environment where the same important information, may and many times is even more important. A PTT whose PDN needs remote data links can use ROSE and determine where the link is coming from for billing or other such needs. A group of PDN's or PTT's who want to

~~229 set up a "regional packet radio network which crosses national boundaries can, with ROSE, determine where the access points to the network are and bill or control accordingly.~~

The other available packet radio networking systems do not have the ability to determine rapidly this information Emergency networking with ROSE will allow a EOC or Civil Defence site determine on to the incoming connect where the call coming initiate first procedures is from and even while the first information frames are starting to pour in. For certain types of emergency operations it is desirable to be able to set up links on which connections can only be made one way, example being an emergency in a given area, you want to be able to connect out but you want to limit connections inbound, ROSE has the ability to do one way links, also has the ability it now to bar calls to a area except call signs only. by certain

I hope this gives you a idea of the CEAMNET up and what and how it was set parts of it have done for the community in Central America, in particular Costa packet radio Rica.It is my personal opinion that can fill in the void that is left in areas where wire line communications are lacking, or need to be extended, it can be used to provide reliable means of communications to remote communities, and extend information and technology to areas there previously this was impossible-OZ

~~230 Experience With Low Cost Digital Communications (or Networking -in and around the University of Yaounde)~~

~~William S. Shu Department of Computer Science, University of Yaounde, B.P. 812, Yaounde, CAMEROON.~~

~~August 1992~~

~~1. INTRODUCTION.~~

Simply looking at current "low cost digital communications" in the University of Yaounde would quickly reduce to listing equipment installed or personal computers linked together. attention - This is not interesting as the equipment do not generally nerd any particular computer except, perhaps, maintenance services - once bought and installed. However, networking is envisaged and so it is worth looking at communication practices around

~~Yaounde as well.~~

~~Computer networks is the type of digital communication I am more familar with in Yaounde.~~

I will therefore broadly discuss some computer networking issues in Yaounde, the CAMPAC network which is the backbone for Cameroon's national and international computer communications, and present relevant details of a proposed campus-wide network for the University of Yaounde.

~~2. IMPRESSIONS ON NETWORKING AROUND YAOUNDE~~

~~My general impression about computer networking in establishments within the city of~~

~~231 Yaound6 is that a few have local networks in place while others have yet-to-be-implemented projects. Those which have (or desire to have) computer networks do so for two main reasons:~~

~~To link micro-computers together within a building, service or site. This generally~~

~~cuts down costs: r single copy of program for many users saves disk space and~~

~~there is greater control over those who access data/software.~~

~~To ease remote processing/data transfers among the branches of an institution and/or~~

~~its head office. Typically, the head office and branches are in different cities. Fast~~

~~responses over secure lines are expected. Installation and running costs do not seem~~

~~a concern (though there is not much choice between CAMPAC and ... CAMPAC).~~

However$ it would seem that management sometimes confuses the requirements for a multi- user, time-sharing system with that for a computer [local area] network. Their decision may be swayed in favour of a network because of a desire to use micro-processors already existing in the company or an inclination to install a system similar to that at Head Office. I must

~~stress that such judgements may be found where computers are used simply as a "tools" rather~~

~~than to sell/provide computer services. In the latter, the decision for a network is studied~~

~~more carefully. Uninformed decisions may have alarming results as the following illustrates.~~

~~A certain commercial institution in Yaoundi installed a token-ring local area network to link~~

~~micro-computers (PS2/386's, 320 Mb). Its applications also needed rapid connections and~~

~~good response times with its agencies in Douala. Shortly after installation, the network was~~

abandoned (before the full quota of micros was obtained) in favour of a multi-user UNIX*

~~UNIX is the trademark of Bell Laboratories.~~

~~232 system. The network system needs intelligent terminals (16 micro-computers costing about~~

7000 US dollars each) and megabytes of disk space for networking purposes, coupled with a high running cost. The UNIX system uses dumb terminals (costing about 200 US dollars each) with no corresponding disk space requirements. Using UNIX, response times and operational costs were greatly improved; the cost of shelving the token ring system was insignificant compared to the profits made.

~~3. THE CAMPAC NETWORK.~~

~~The CAMPAC (CAMeroon PACket) network is the backbone of computer communications, within and beyond Cameroon. CAMPAC became operational in 1984 and is currently under~~

~~INTELCAM, a division of the Ministry of Posts and Telecommunications (PTT). It is an~~

X.25 packet switched network seryiEe that conforms to the international requirements defined by CCITT and the International Standards Organisation (ISO). Thus, it also handles communications across national boundaries directly.

~~3.1. CAMPAC Structure.~~

~~The structure of the CAMPAC system is given in [Figure 1]. It consists of:~~

~~3 switching nodes: 1 each in Douala, Yaound6, and Garoua; another is proposed for~~

~~Bafoussam.~~

~~2 concentrators: 1 each in Douala and Yaound6.~~

~~Dual links between nodes and duplication of equipment for rapid communications.~~

~~1 network management centre in Douala.~~

~~N-has a satellite link to the rest of the world through the "Noeud de Transit~~

~~233 International (NTI) de Paris", an international gateway for packet-switched~~

~~communications networks in France.~~

~~- Access to the network may be done through special [highspeed] lines, through~~

~~switched telephone systems or modems.~~

~~3.2. Access/Services of CAMPAC~~

The CAMPAC network uses full duplex virtual circuits with data rates adjustable to that for the receiving device. The X.25 high-speed lines operate at 2400 to 9600 bps. For economic reasons, these lines are currently operated at 9600 bps while access through switched telephone networks operate at 1200 bps. Three main user services are currently provided:

~~Direct connection via the special lines. These are mostly leased to large companies~~

~~and banking institutions.~~

~~Access via telephone lines and modems.~~

~~Facilities for users, notably in academic institutions, to access databases and other~~

~~utilities (e.g. NUNTML in France) all over the world.~~

~~About 100 subscribers are connected directly via the special lines while another 150 use the~~

~~other services. INTELCAM workers hold that the number of subscribers ' e increasing as~~

~~more people become aware of their services. Furthermore, the telephone network has of late~~

~~been quite reliable. Users do not generally have problems in using CAMPAC once it is~~

~~installed. Most difficulties met stem from poor knowledge of their modems and/or computers;~~

~~occasionally, the underlying telephone service (which is not under INTELCAM of course!)~~

~~may malfunction.~~

~~234 4. UNIVERSITY OF YAOUNDE.~~

The University of Yaound6 has no campus wide network services, though there are plans to install one. There may be the odd local area network (LAN) which, as mentioned earlier, may require practically no attention once installed. Currently, there is a special line linking the Computer Centre of the University to CAMPAC. Electronic communications can thus be done. but in a rather convoluted manner. A user logs on - via CAMPAC and NTI - as a remote terminal/user to a machine in, say, the USA. He can then use all the facilities available on that machine, including network communications. He can also download files/data from the machine to those in the Computer Centre.

~~For example, our head of Department (Computer Science) has a login account in "Institut de~~

Recherche en Informatique et Sysumes A16atoires" (IRISA), France. The rest of the department, subject to his consent, may "borrow" his address to receive e-mail from other countries. Currently, the connection time for such accesses is about 10 to 15 minutes per day.

~~However, things are expected to change in the near future; first within the Computer Centre and then within the university.~~

~~4.1. The "Projet FAC"~~

The "Projet FAC" [Go9l] was conceived to provide the University of Yaoundd and its professional schools (its "grandes dcoles") with modern computer facilities to meet teaching, research and administrative needs. The proposed solution for the whole university - as opposed to specific departmental requirements - is based on a campus-wide. ETHERNET network that links (at 10 Mbps) the various schools, faculties, departments and buildings. For

~~235 schools further away from the main university campus, separate lines (9600 bps) are envisaged. A link through CAMPAC provides access to other networks all over the world.~~

~~See also [Figure 2] which is obtained from [Go9l].~~

Stations linked to the network will use standard protocols such as TCP/IP and NFS. The hardware are personal computers (PCs) under MS-DOS' or UNIX workstations. Decisions made are based on the need to make full use of existing resources (e.g. better utilisation of

~~available PCs), minimise duplication of resources by uniformly treating all departments, and~~

~~to ease the acquisition and use of packages and materials popular in research communities.~~

~~Some of the technical details for the network follow.~~

~~a) The campus-wide network: The technologies/protocols used are:~~

~~Medium access control as given by the IEEE 802.3 (or ISO 8802.3) standard.~~

~~Optical fibre technology (optical fibres, repeaters and star couplers) between~~

~~buildings.~~

~~Coax cables or twisted pair within buildings.~~

~~If possible, assorted communications structures supporting different technologies (e.g.~~

~~ETHERNET, FDDI); multi-mode optical fibres (62.5/125)~~

~~b) PC Needs:~~

~~about 50 ETHERNET cards for that many PCs.~~

~~TCP/IP etc. for PCs.~~

~~' MS-DOS is a trademark of Microsoft Corporation.~~

~~236 c) UNIX Machine needs:~~

~~ISO 8802.3, X.25, HDLG and X.21 interfaces.~~

~~TCPAP including FT?, TELNET, RLOGIN, SMTP, and SENDMAIL for all the~~

~~network interfaces considered.~~

~~Network File Server (NFS).~~

~~S. CONCLUDING REMARK.~~

It would seem that the full potential of computer communications is not exploited by establishments that have computer networks in Yaoundd. Also, a number of establishments tend to have network projects under study. Fortunately, the need for computer networks is increasingly aparent. Its implementation, however, is seriously constrained by economic and administrative factors. In academic communities the need is overwhelming both in teaching and research. It is also useful in side-stepping tedious administrative procedures ij sending urgent information. (For instance, in sending a FAX, I go through two secretaries and two signatures, but sending a-n}ail needs only the e-mail address!) However, efforts are being made at the University of Yaound so that network services come to stay. I hope that establishing it would be sooner, not later.

~~6. REFERENCE~~

~~[Go9l] Goff, H. Le, "Projet FAC: Universiti de Yaound6 et dtablissements associes (ENSP,~~

~~ENS, ...); Relevd de propositions et de recommandations suite A la mission effectuee~~

~~a Yaounde du 28 mai 1991 au 01 juin 1991", (IRISA-INRIA, June 1991).~~

~~237 Europe Japan (DDX-P)~~

~~Gabon USA (TELENET/THYMNET) -- NTI CBte d'Ivoire - Tunisie Key: LS E special lines T a Access point to switched telephones Access to NTI 9600, X75~~

~~BAF OU SS AM~~

~~GAROUA 46000 bps, X25~~

~~I 9600 bps, X25 Network f- { X 25 X25 X25 LS access Manage., -- switch LS data (9600 bps) Centre switch 9600 bps, X25~~

~~Conc entT,ator PAD PAD Concentrator in Ddu'ala in Yaound6~~

~~Access Access via Access through LS Switched tel. through LS service~~

~~DOUALA YAOUNDE~~

~~FIGURE 1: The CAMPAC Network~~

~~238 PC~~

~~V mCULTI des SCIENCES tz~~

~~CUSS COAXIAL CABLE OPTICAL STAR COUPLER FIBRE OPTIC LINK WORK STATION ENSP GENERAL ADMIN. MACHINE RESEARCH MACHINE MICRO-PROCESSOR~~

~~FIGURE 2: Campus-wide network of the University~~

~~239 240 Technical Background Papers~~

241 242 Digital Communications via HF radio ("High Frequency" - between 3 and 30 MHz) has proven to be more difficult than similar communications at higher frequencies, either VHF or UHF ("Very High Frequency" - 30 to 300 MHz; "Ultra High Frequency" - 300 to 1000 MHz). This section begins with a technical introduction and overview of digital radio systems followed by papers describing some of the technical improvements currently underway to make HF packet radio systems more reliable and capable of faster.transmission speeds.

~~243 244 ERROR-CORRECTING RADIO SYSTEMS~~

~~A Discussion of SITOR and Packet Radio~~

~~presented for~~

~~VOLUNTEERS IN TECHNICAL ASSISTANCE Arlington, Virginia May 1991~~

245 A SMALL HISTORY OF RECORD COMMUNICATIONS This background material was originally prepared for a series of seminars in telecommunications presented by the writer to various Bell Operating Companies for Siemens Data Switching Systems, Inc. Much of the following historical information was drawn from the official archives and files of the International Telecommunications Union (ITin, during visits to its World Headquarters at Geneva.

The TTU is an agency of the United Nations, founded in Paris in 1865. As a function of its technical and non-political nature, the TTU attempts to coordinate telecommunications operations and protocols between the nations of the world. The TTU Archives record all of the significant advances, developments and patents in the communications arts and sciences during the last 100 years.

These archives are generally recognized as non-political in nature, and are usually accepted as an authoritative source of information on the history of the telecommunications industry. Because some of the information found in the TTU Archives may not agree with what is taught in the American school systems and possibly seem controversial, the following statements drawn from the TTU Archives are presented without further comment or editorial opinion by the writer. The reader can judge the merits of the arguments presented herein.

~~IN THE BEGINNING .. .~~

The first step in the direction of plain language telegraphy came in 1855, 10 years before the Paris Conference, when David E. Hughes was granted a patent for a new telegraph. His apparatus was used mainly in Europe, as the patent position in the U.S.A. prevented him from exploiting his invention here. Hughes' French partner, Gustave Froment, brought the new telegraph to a high state of perfection; however, its mechanical parts were complex and often broke down. Even so, it was a great advance over the Morse system which could send only 25 words per minute - even the early types of the Hughes telegraph achieved between 40 and 45 words per minute.

The principle of the Hughes telegraph was simple. A continually rotating wheel had 28 letters of the alphabet and other signs on it. A clutch mechanism actuated by an electromagnet brought the wheel momentarily to rest when the desired letter was over a moving strip of paper. Ink was continually fed to the type wheel from a roller, and thus a message in plain language could be spelled out straight onto the receiving paper. For sending a message, the operator had a piano keyboard in front of him, which relayed the electric impulses to the receiving post. These pulses stopped the type wheel at the correct instant of time. Over the decades, the modern automatic telegraph printer has grown from this basic principle, but only after many improvements in layout, design and detail.

246 The next step forward in telegraphy came from Emile Baudot, an officer in the French Telegraph Service. In 1874, Baudot-developed the rotary distributor for telegraphy, and introduced the five-unit code using five bits to define each character, with a "start" bit and a "stop" bit to identify the beginning and end of each character. For each letter, therefore, five impulses were sent over the telegraph line, which actuated a set of five magnets at the receiving end. Baudot combined the use of the five-unit code with the time division multiplex system, thus allowing several telegraph communications to be transmitted over the same circuit.

The sending operator used a special keyboard of five keys; on the receiving side, the strip of paper passed through an electrical printer, printing out the telegram in clear language. In modern machines using these principles a speed of 60 words per minute, or one six-letter word per second, is allowed for in the design of the telegraph, it being considered the maximum speed at which operators can work economically for a long time. Baudot's system was first introduced in 1874, and officially adopted by the French Telegraph Service in 1877.

Donald Murray, a New Zealand farmer turned journalist, inventor of the Murray Multiplex System.(another five-unit code system) made a valuable contribution to telegraphy by rationalizing the allocation of the combinations of the of the five-unit code to the characters of the alphabet on the basis of frequency of occurrence. His arrangement of the code, in which the most-frequently-used letters of the alphabet are represented by the smallest number of holes in the tape, has since become the five-unit "Baudot" code in today's standard practice.

Although the United States of America had been linked to Europe by the two Atlantic cables of 1866, private American telegraph companies were not represented during the first conferences of the International Union. The first to come was a representative of the Western Union company; he went to the Berlin Conference of 1885. But the ever-increasing demands on telegraph services, in the United States as well as in Europe, led to many important technical improvements of the American telegraph system. One of the most important of these was the sending of more than a single telegraphic message over the same wire. It was Thomas Alva Edison (1847-1931), who in 1874 advanced this technique more than anyone else.

There can be little doubt that international cooperation existed on the technical level of telegraphy as early as it did on the administrative side. The results of the discussiogs of the technical committees of the International Telegraph Union found their way across the Atlantic to the American telegraph engineers.

The first to use a single telegraph wire for the sending of two messages, one in each direction, was a Dr. Gind in Vienna in 1853. Without going into the technical details of his "duplex" circuit, it meant the incorporation of a balance filter in the circuit, thus making it unaffected by the signals sent out, but fully responsive to those coming in. This led to such

247 economic advantages that many others carried on his work and improved it on both sides of the Atlantic. Artificial lines, Differential Duplex circuits, and Bridge Duplex circuits were all steps in this development.

Edison, who had himself earned his living as a telegraph operator since the age of 15 - he was still only 18 years old when the Paris Conference met in 1865 - naturally turned his superb inventive brain to the improvement of the telegraph services of his time. He took out patents for duplex circuits and in 1874 he invented the quadruplex circuit, making it possible for four messages, two in each direction, to be simultaneously transmitted over a single telegraph wire.

This trend of sending more and more messages over the same electrical conductor has continued ever since; in telegraphy it found its perfection in the multiplex system, and in telephony we can now transmit thousands of simultaneous telephone conversations over the same coaxial cable. If we think back for a moment to the early pioneers of telegraphy, using a single separate wire for each letter of the alphabet, and compare that with the thousands of simultaneous telephone conversations, we can see at a glance how tremendously the communications engineer has contributed to the progress of civilization.

Undoubtedly, both the duplex circuit and the Hughes' telegraph were discussed at the Paris Conference of 1865. To some they might have appeared as much science fiction as Jules Verne's cannon shot to the Moon. But to others, the new ideas were a stimulus to improve the international telegraphy service both on the technical side and from an administrative point of view. The Paris Conference was certainly an outstanding event and had set the pattern for international cooperation for the next hundred years. As the delegates travelled back to their respective homes, by railway or by horse, they could certainly be proud of the work they had achieved.

~~ENTER THE WMLESS ...~~

The story of radio begins perhaps with Joseph Henry (1797-1878), an American physicist, who discovered in 1842 that electrical discharges were oscillating. A gigantic step forward was that due to James Clerk Maxwell (1831-1879), a Scottish physicist and one of the great mathematical geniuses of the 19th century. His treatise on "Electricity and Magnetism" was read as a paper to the Royal Society in 1864, and published in fully developed form in 1873. It has been called "one of the most splendid monuments ever raised by the genius of a single individual". By purely mathematical reasoning, Maxwell showed that all electrical and magnetic phenomena could be reduced to stresses and motions in a medium, which he called the ether. Today we know that this "imponderable electrical medium" does not exist in reality, any more than the geographer's equator, or the statistician's average man.

~~Yet the concept of an ether helped greatly, and allowed Maxwell to put forward his theory that the velocity of electric waves in air should be equal to that of the velocity of light~~

248 waves, both being the same kind of waves, merely differing in wave length. This we know today to be an elemental truth, yet to-Maxwell must go the honor of having first shown it to us in pure mathematical form.

In 1857 Feddersen demonstrated that if an electrical condenser is discharged into a conductor, oscillations are set up which give rise to intermittent spark phenomena. Twenty- one years later, in 1878, David Edward Hughes (18311900), an Anglo-American physicist, made another important discovery in the prehistory of radio and its essential components; he found that a loose contact in a circuit containing a battery and a telephone receiver (invented Bell by in 1876) would give rise to sounds in the receiver which corresponded to those that had impinged upon the diaphragm of the mouthpiece. Hughes' "microphone" consisted of a carbon rod resting in grooves of two carbon blocks; from it developed many of the early carbon microphones of both telephone and radio. In 1883, George Francis Fitzgerald (1851-1901), an Irish physicist, suggest a method by which electromagnetic waves might be produced by the discharge of a condenser.

Next we must turn to Heinrich Rudolph Hertz (1857-1894), the famous German physicist, who was the first to create, detect and measure electromagnetic waves, and thereby experimentally confirmed Maxwell's theory of "ether" waves. In his experiments he showed that these waves were capable of reflection, refraction, polarization, diffraction and interference. They corresponded precisely in their behavior to waves of light. Hertz produced his waves, soon to be called by others "Hertzian waves" from the sparks of an induction coil, and in order to study some of their properties he employed a zinc mirror. He described one of his experiments, in 1888-89, as follows:

"The height of the ,(parabolic) mirror was thus 2 meters, the width of its aperture 1.2 meters and its depth 0.7 meter. The primary oscillator was fixed in the middle of the focal line. The wires which conducted the discharge were led through the mirror; the induction coil and the cells were accordingly placed_ behind the mirror so as to be out of the way. If we now investigate the neighborhood of the oscillator with our conductors, we find that there is no . action behind the mirror or at either side of it; but in the direction of the optical axis of the mirror the sparks can be perceived up to a distance of 5-6 meters". The half-wave length of this experiment was about 30 cm.

Hertz died at the early age of 37, and once said " ... these are the ultimate problems of physical science, they icy summits of its loftiest range. Shall we ever be permitted to set foot upon one of these summits? ... We know not: but we have found a starting point for further attempts which is a stage higher than any used before". There could be no better epitaph to one of the greatest of inventors of wireless telegraphy.

But Hertz's experiments were half a century in advance of his time, and belong to the field we now call microwave optics. Many repeated them and extended them, typical amongst them being Edouard Sarasin (1843-1917) and Lucine de la Rive (1834-1924) at Geneva, Antonio Giorgio Garbasso (1871-1933) and Emil Aschkinass (1873-1909) at Berlin, Jagadis

249 Chunder Bose (1858-1937) at Calcutta and Augusto Righi (1850-1920) at Bologna. We shall see in a moment that only Righi had an indirect influence on the technology of radio, as a young Italian by the name of Guglielmo Marconi (1874-1937) was stimulated by his books and lectures.

The next most important event was a lecture given to the Royal Institution in London by Oliver Joseph Lodge (1851-1940) on June 1, 1894; it was called "The Work of Hertz and Some of his Successors% it was widely reported at the time, and was to have far-reaching consequences. Lodge, Professor of Physics at the new University of Liverpool, had himself worked extensively in the field of electromagnetic waves and was the first to comment on the phenomenon of resonance or tuning. In 1898 he took out a patent on an adjustable inductance coil in the antenna circuit of a wireless transmitter or receiver, or in both, in order to tune one with the other. This patent won him a high place in the history of wireless. His sharply tuned resonance circuits were a substantial advance over Hertz's relatively primitive arrangements, but Lodge himself, like Hertz with academic commitments, had not the time to develop his ideas on wireless telegraphy and never made an attempt to transmit an intelligent signals with electromagnetic waves.

Alexander Stepanovitch Popoff (1859-1906) was one of the many who read Lodge's lecture and was inspired by it. Popoff was in 1895 lecturer in physics at the Russian Imperial Navy's Torpedo School at Kronstadt near St. Petersburg. He experimented with Branly coherers (see below), set up a receiver with a protruding wire in 1895, and read a paper about it, "On the Relation of Metallic Powders to Electric Oscillations", at the meeting of the Russian Physico-Chemical Society of April 25 (May 7, New Style), 1895. By means of this equipment, Popoff could register electrical disturbances, including atmospheric ones; and in July of the same year a similar instrument with an ink recorder was installed at the Meteorological Observatory of the Institute of Forestry at St. Petersburg.

A more complete description of his experiments appeared in January 1896, and on March 12 (24 N.S.) of the same year he gave a further demonstration before the same society. Only a very brief minute appeared at the time in the records of the society: "A.S. Popoff shows instruments for the lecture demonstration of the experiments of Hertz. A description of their design is already in the Zh. R.F.Kh. Obshchestva." At that meeting the words "Heinrich Hertz" were transmitted by wireless telegraphy in Morse code and similarly received before a distinguished scientific audience. To the Soviet Union, Popoff has become the inventor of the radio, May 7 being celebrated each year as "Radio Day".

In view of the date of Marconi's first patent (it was applied for on June 2, 1896, the specification was completed on March 2,1897, and it was accepted on July 2,1897) there has in recent years been much discussion about the priority of the invention of radio. We have already seen a similar case in the invention of the telephone, when only a few hours separated the filing of a caveat and a patent application by Gray and Bell; in Germany, Reis is considered "Der Erfinder des Telephons" as it is recorded on his memorial statue.

250 Most historians of science and technology take the view nowadays that, given the right preliminary stimulus - here the work of Hertz and the lecture by Lodge - and the availability of the right materials and equipment, then a given invention might well be made simultaneously by more than one research worker. Once more looking back, but to the history of the early telegraph, we found that it was not only the act of invention that counted in technology, but also the successful communication of the idea, so that it could be applied and brought to fruition. There can be do doubt that in the case of radio Marconi invented a system of highly successful wireless telegraphy, and that he personally inspired and supervised its application until it spanned the world. This must be considered as ample justification for his ward, in the year of 1909, of the Nobel Prize for Physics. He shared it with Carl Braun (1850-1918) "in recognition of their contributions to the development of wireless telegraphy," to quote the official citation of the Nobel Foundation at the time. Braun will always be remembered for his development of the cathode ray tube, the direct ancestor of the Kinescope and the television tube, quite apart from his other work in radio.

Marconi, at the age of twenty in 1894, was well acquainted with the work of Hertz, Branly, Lodge and Righi. As the son of a rich country gentleman, his education had been private, and his main interest lay in science and music. He began experimenting in the spring of 1895 on his father's estate at the Villa Grifone, near Pontecchio, Bologna. The idea had occurred to him that Hertzian waves might be the basis for a means of communication, signalling with them the dots and dashes of the Morse alphabet. Once he had convinced his father of the practical nature of his ambition, he received from him all the financial support he needed. .

In his first experiments, he used an ordinary spark induction coil and homemade coherers of the Branly type (see below). To turn the discharge on and off, he placed a telegraphy key in the primary circuit of the induction coil, and thus produced short or long trains of sparks. He could soon detect these dots and dashes all across the room, and in the summer of 1895 he moved out into the garden. In order to increase the performance and range of his transmitter, he followed previous experiments by others: he attached to one end of his transmitter circuit an elevated metallic object, the antenna, and to the other end a metal plate buried in the ground. He could now signal across the whole garden, and he soon found that there was a direct relationship between the height of the antenna and the distance of transmission.

Marconi made important improvements in the components of his system, particularly in the design of his coherers. He also put a relay in series with it, which actuated the tapper of his coherers, and worked a telegraphic printing instrument to record the received signals.

Here is perhaps the best place to pay tribute to Edouard Branly (1844-1940), a great French physicist, officially described in France as "Inventeur de la Telegraphic Electrique sans Fil". His great contribution was the discovery of the coherer, that small fragile glass tube, looking like a thermometer, but filled with metal powder. He demonstrated it before the French Academy of Science, to which he was later elected, in 1891.

251 Branly found that electromagnetic waves, produced as much as 25 meters away from it, caused the individual metal particles in his coherer, first iron and later nickel and silver, to cohere and thus allow the passage of a current through them. A galvanometer was the instrument he used to show this effect; Marconi improved this greatly by using a telegraph printer. But the metal particles had to be separated again, and therefore an electric tapper, a tiny hammer precisely like that used in any electrical bell was added to the coherer. When it struck the glass tube it decohered the particles again, and thus stopped the current from the batteries.

Each successive impulse reaching Marion's antenna produced the same phenomena in the coherer, first the coherence of the particles, then their decoherence, and hence the recording of the dots and dashes. Marconi used tightly fitting silver plugs in his glass tube, which he evacuated and sealed and thus the coherer became the first of many sensitive devices to receive wireless telegraphy. Before Marconi left Italy, to continue his work in England, he had reached a transmission distance of the order of 1 kilometer.

Such then is the story of the many inventors of wireless telegraphy, working with each other's equipment, adding new ideas and new improvements to them. It was a patient, persistent and often discouraging inquiry into natural laws and in these initial stages it was indeed only animated by the love of knowledge.

~~SOME FUNDANmNTAL CONCEPTS RTTY~~

The term "RTTY" does not define any single data code or speed. Contrary to some information published in recent radio equipment manuals, "RTTY" means only one thing - "radioteletypewriter", or "radioteleprinter service". It does not mean "radioteletype" ! The term "teletype" is a registered trade mark of the Teletype Corporation, and strictly speaking, should be used only when referring to that company's products.

"RTTY" does not mean Baudot - ASCII code is also a form of "RTTY". SITOR modes and packet radio are also forms of "RTTY". Think of human communication; all humans share a common communication mode: speech. The languages change from place to place. Baudot-Mur ay code and ASCII and SITOR codes are just different languages used in the same mode - the difference between French and Spanish and Italian and Portuguese! If you extend this rationale, SITOR, and packet radio are additional digital languages within the same mode, but with flavors all their own.

~~252 BAUDOT-ANRRAY~~

The Baudot-Murray code dates. from about 1874, when F-mile Baudot, a Lieutenant in the French Telegraph Service, developed the Baudot distributor for telegraphy, using five data bits to define each character, with a "start" bit and a "stop" bit to identify the beginning and end of each character. With five bit positions, the total number of possible combinations equals 25, or 32. Our language has 26 letters (upper case only!), 10 numerical digits, and 9 common punctuation characters ... a total of 45 alphanumeric characters. The Baudot- Murray codes resolve this seeming paradox by the same stunt used on the classical typewriter - shift to upper case and you now have the numerical digits, the punctuation marks, and a bunch of other useful things like parentheses, dollar sign, number sign, "at" sign, asterisk, percent sign, ampersand, underscore. This produces a total of 64 possibilities. That's fine if you never send anything but business letters or need to talk to computers!

Different versions of the same Baudot-Murray code evolved, causing confusion in international traffic. Western Union, Bell, the Weather Bureau, the Armed Forces, the F.A.A., the Associated Press, all had their own versions of the code. The situation became problematic when subscribers couldn't get together on the same keyboard configurations. The International Telex Network used yet another version of the Baudot-Murray code called International Telegraph Alphabet Number 2, the version called out for the U.S. amateurs in Part 97 of the Commission's rules.

~~Although still the most widely used code in the world (the International Telex Network is still the largest of all record communications networks), Baudot-Murray code has two serious drawbacks:~~

~~There is no parity or inherent method of validating transmission integrity; a machine has no way of telling if an error has occurred~~

~~The code is a sequential one, meaning that a particular control character defines the following series of characters for a period of time until a new control character is recognized.~~

The two control characters that identify the functional configurations in Baudot-Murray code are "LETTERS" and "FIGURES". Those of you who operated RM in the days of electromechanical teleprinters and Teletype machines know what happened when the data was kicked into upper case by a noise hit, and how frustrating it was to have lines of unintelligible stuff.

~~DATA CODES~~

~~Baudot-Murray code was fine stuff back in the days when the game was communication between humans. As technology developed and machines had to start communicating with~~

253 other machines without involving humans, better and more efficient codes had to be developed for transmitting information, so that the machines themselves could evaluate the received information and request repeats as needed in the event of errors being received.

As data processing began to evolve and data communication codes were developing, the data processing systems used their own codes, one of the first of these being "BCD", Binary Coded Decimal. BCD was used for internal calculations inside a data processing device. BCD has no alpha characters, only numbers, and thus was unsuitable as a communications code for use by humans. ,

~~BCDIC~~

Binary Coded Decimal Interchange Code was developed when the data processing systems had to communicate with humans in printed characters on a printing device. This code was fine for communication with humans, but was unsuitable for machine-to-machine communications because it lacked any form of parity or error-checking capabilities. So, BCDIC had the same problem as did Baudot - a machine couldn't tell when an error occurred in transmission.

~~EBCD~~

Extended Binary-Coded Decimal solved the parity problem when it was developed for the IBM Selectric typewriters. The Selectric principle allowed the typing and printing of hard copy while at the same time, generating a unique code suitable for transmission over a communications facility. Also known in the field as MC (Paper Tape Transmission Code), this code used six information bits and a parity bit which permitted the receiving system to determine if an error had occurred in a specific character. But, Extended BCD is a sequential code with upper case and lower case characters. So, while parity could detect errors and provoke repeats, the sequential nature of the code made its efficiency less than desirable.

~~EBCDIC~~

Extended Binary-Coded Decimal Interchange Code was developed in 1962 and, with its 8 bits, was capable of transmitting 256 characters. But for communications, EBCDIC might be deemed a backward step, because it had no parity capability. Some users who don't need all 256 characters have redefined the code using EBCDIC as a base and identifying particular characters with odd or even parity bits. Although there are occasional compatibility problems because of parity definitions varying between users, EBCDIC is still widely used in data processing communications systems, and is the internal data code in many IBM main

~~254 frame computers and computing systems.~~

~~ASCII~~

ASCII was published as a standard by ANSI, the American National Standards Institute around 1963, in order to achieve some degree of compatibility for the newborn data communications field. Contrary to the beliefs held by some computerists, ASCII is an extension of the eight-bit code developed,nearly fifty years ago and used for years in the Bell System TWX network. ASCII is also known as International Telegraph Alphabet Number 5 and is standardized for international traffic at various data rates.

ASCII uses seven bits to define each character and gives us a maximum of 20, or 128 possible combinations. We now have the ability to transmit upper-case and lower-case letters, all punctuation marks, 10 digits, and a variety of control codes such as "start of text", "end of text", "horizontal tab", "vertical tab", "form feed", "backspace", and many other control functions previously unavailable in Baudot-Murray code.

ASCII is frequently thought of as an eight-bit code; the eighth bit is reserved for a parity function, a form of error detection. Many ASCII systems do not require the eighth bit for parity and operate well without it.

~~OTHER CODES~~

Several other data codes are in use today, some of them dating back decades. The Hollerith Code, developed more that 80 years ago, is still used commercially and is generally associated with punched-card systems. There are others such as Jaquard Code, Moore ARQ code used in time division multiplex, six-bit Transcode, Syntoc code, AP code, and others too numerous to list here.

After many changes in radio rules, the communications operator can use many or all of these various codes, as long as they can be documented. This possibility becomes more interesting when one realizes that one of the main advantages of packet radio is that the packet network concept is really transparent to the user - the network doesn't care what code or data speed the user's terminal is sending to the network! The relationship between the packet controller (the box that makes the packets out of what you send from your terminal) and your keyboard/terminal/computer system is flexible and can meet almost any user's need, now and in the future!

~~255 ASYNCHRONOUS VERSUS SYNCHRONOUS TRANSMISSION~~

Since the beginning of electromechanical telegraphy and teleprinting, maintenance of the proper timing relationship between the sender and the receiver has been resolved by mechanical synchronization methods. These were based on the basic ideas of "start-stop" telegraphy. The original Baudot rotary distributor permitted the sending distributor to tell the receiving distributor when the rotary contact was at a reference point in the rotation of the motor shaft. The distributor created a start and stop pulse at the beginning and end of each character.

In conventional teleprinter operation, each character has carried its own sync information in the form of the traditional "start" element at the beginning of each character, and the "stop" bit at_ the end of the character. In some systems, the "stop" element is one, one-and-one- half, or two bits. The result is reduced efficiency - each character will have a fixed amount of "overhead". These housekeeping bits carry no user information. They exist only to keep the sending and receiving machines in sync. This reduced efficiency has kept a lid on higher data rates and has been a block in the constant search for faster ways of sending more information for more people.

As data processing requirements became more demanding, newer forms of synchronization evolved. A method was developed which applied a single synchronizing string to an entire string of characters, rather than include sync information in each character. The "overhead", or housekeeping bits previously inserted into each character for sync purposes were eliminated; higher data rates became practical, without sacrificing data integrity. This newer method, called "synchronous" transmission, is used in higher-speed and specialized forms of transmission, including both SITOR and packet radio. Even though your terminal keyboard sends your packet controller (or SITOR system) asynchronous data, with start and stop bits in each character, packet controllers and SITOR converters strip the start and stop bits from your characters.

~~256 SITOR - ERROR-CORRECTING TELEPRINTER OVER RADIO~~

~~In this paper, we'll use the term "SITOR" as a generic reference to various applications of the same error-correcting radioteleprinter system.~~

~~SITOR - DEFINITIONS AND STANDARDS~~

STTOR (Simplex Telex Over Radio) is the name applied to a method of error detection and correction defined in CCIR Recommendation 476, for Direct- Printing Telegraph Equipment in the Maritime Mobile Service.

AMTOR (AMateur Teleprinting Over Radio) is the name derived by the Amateur Radio fraternity to describe the STTOR system as modified and applied in amateur radio operation. AMTOR is authorized by the FCC for the Amateur Radio Service in Part 97.69 of the Commission's Rules, and must comply with CCIR Recommendations 476-2 and 476-3.

~~Our references to "ARQ" will encompass both SITOR and AMTOR.~~

~~GENERAL SITOR CONCEPTS~~

When high-frequency radio is used for teleprinter telegraphy, the mutilation of data bits is normally much worse than with land-based telegraph circuits. BF radio is subject to severe fading and distortion, especially in times of high sunspot activity. Because of its high error rate and general unreliability, its use is avoided insofar as possible for the transmission of computer data. However, BF radio continues to be widely used by high-seas vessels in ship- to-shore teleprinter and telex links, and in many isolated areas.

BF radio links can frequently have an error rate (before any correction) of one character in every thousand - sometimes much worse. The mutilation rates can rise as high as one character in forty, or even as high as one in four on bad links and at certain bad points in time.

The transmission of information over BF radio can be protected against fading, noise and other disturbances by use of an "ARQ" system (Automatic Request). SITOR equipment, also known as "TOR" (Teleprinting On Radio) prevents messages from getting wholly or partially lost.

~~257 FIRST PRINCIPLES~~

~~The use of a special transmission code permits the receiving system to determine if a character was garbled or mutilated during transmission.~~

There are 32 recognizable characters in the Baudot-Murray teleprinter and telex systems. These are normally transmitted by 32 combinations of five data bits. If five data bits are used, any bit error will transform one character into another and the error will pass undetected. ,

~~THE SITOR CODE~~

SITOR uses seven data bits, giving 128 possible character combinations, of which only 32 are recognizable printing characters. Careful choice of which 32 are used minimizes the possibility of an undetected error.

Only those combinations with three zeros and four ones are used, making it easy to check for errors at the receiving end. There are 35 such combinations or codes, and so the three extra bit sequences are available for control purposes. These three control codes are:

~~The RQ character signals the transmitting station to "Please repeat" The idle Beta character The idle Alpha character~~

The CS 1 and CS2 codes and CS3, to be explained shortly, are also seven-bit characters from the same set. Because these control codes are always sent only in the reverse direction, they are never confused with the others. Note that this code is designed to translate easily to and from the Baudot-Murray code.

When the receiving system decides that the received character is mutilated, the output of the STTOR system is blocked so that the mutilated character is not passed on to the teleprinter or terminal. At the same time, a repetition request (RQ) is automatically returned to the station that sent the character.

The repetition request is made with the aid of a specially-designed transmit and receive system. When such a request for repetition is received at the transmitting station, it starts a repetition cycle to the effect that a group of last-transmitted characters (including the mutilated one) is retransmitted.

~~258 SITOR PROTOCOL - MODE A - "ARQ"~~

In the STTOR mode (A) or ARQ, Station A sends burst of synchronous frequency-shifted data containing three seven-bit characters. Station B, in response, sends the acknowledgement signals in the reverse direction as a single character. How, then, can ARQ solve the communication problem? The answer lies in encoding the acknowledgement signals differentially using two control characters called CS 1 and CS2.

When Station B is copying correctly, receiving valid characters, Station B replies with CS 1 and CS2 alternately after each block. If a bad block is received, Station B repeats the same control code as the last time. If Station A sends "Please repeat," then Station B repeats the same control code as the last time. Thus Station B's reply is the same for a "Please repeat" block as for an error. It doesn't matter, therefore, if the bad block was a "Please repeat" block.

To prevent confusion, the changeover in direction of transmission is not left to the operators; there could be a misunderstanding if the link fades out just before the expected end of a transmission. There are two ways to signal for the changeover.

~~The sending station can end his transmission with the two-character sequence +? The receiving station operator can press the TRANSMIT button.~~

By either method, the receiving station stops replying with CS 1 or CS2 and instead acknowledges with CS3. Upon receiving this information, the sending station transmits the block "beta-alpha-beta". In response, the receiving station transmits an RQ character, whereupon the transmitting station goes to the receive mode.

Bursts of data from each station are so timed that even if both are transmitting blocks momentarily, each one can still receive one character of the other's block in the position expected to be a control code. This seemingly complicated process does ensure that the changeover proceeds in an orderly manner and cannot go awry, no matter what.

Timing of the various signals is shown in the figure, with some of the possibilities for errors. Note that the two stations do not behave identically in respect to timing. One is called the "Master" station and the other the "Slave", for reasons which will become apparent shortly.

~~SITOR PERFORMANCE~~

Although ARQ, in common with other error-correction systems, almost completely eliminates errors resulting from the radio link, it is worthwhile pausing to see exactly just how good it is. A simple analysis can be made by supposing that the radio link alternates between perfect copy and perfect random noise. With only noise in the receiver, all 128

~~259 seven-bit patterns are likely to be received with 34 of these being acceptable (the RQ character is treated the same as an.error).~~

Thus the chances of a whole block of three characters being accepted by mistake is (34/128)3 or about 1.9%. Therefore, with no signal, the receiving terminal will be idle for 98.1% of the time while the system is asking for repeats and will be displaying garble for 1.9% of the time. This compares with 100% correct copy when the signal is good.

By using the foregoing information, we can calculate the proportion of garble to good copy for various proportions of good signal tb bad. A similar analysis for the reverse path shows that when there is no signal in this direction, 0.8% (1/128) of the message is unwittingly lost into thin air. The combined effect of these factors is shown in the table.

~~The assumption is made that the signals in both directions alternate between perfectly good and perfectly bad.~~

Percentage of Percentage of Number of Time Taken as Time Signal Transmitted Spurious Characters Multiple of is Usable Message Received Printed as 100% Signal Correctly Percentage of Case Transmitted Message Length

~~100.0 100.0 0.0 100.00 90.0 99.9 0.2 1.11 80.0 99.8 0.5 1.25 70.0 99.7 0.8 1.42 60.0 99.5 1.2 1.66 50.0 99.2 1.9 2.00 40.0 98.8 2.8 2.50 30.0 98.2 4.4 3.30 20.0 96.8 7.5 5.00 10.0 93.0 16.9 10.00~~

~~. 5.0 85.2 35.6 20.00 2.0 61.7 91.8 50.00 1.0 22.7 185.5 100.00~~

~~260 SITOR SYNCHRONIZATION~~

Considering that SITOR is a synchronous system with no start and stop bits, the timing at both ends must be stable. Some means must be found to get the two stations in step and to keep then that way over a period of time, even if the two clocks are only slightly different in speed.

The synchronization procedure starts with the first station (the master) sending a special synchronization block repeatedly. The slave station continually shifts in received bits until 21 consecutive bits correspond exactly with the expected synchronization pattern. The slave then starts to reply in the gaps, sending back one of the control codes. The master station, meanwhile, has been shifting in received data bits during the gaps in its transmissions. When it recognizes two consecutive control codes, it stops sending synchronization blocks and changes to sending traffic. In fact, to guard against the possibility of the slave getting the synchronization pattern right, just by chance, the master sends two different synchronization blocks alternately, and the slave must get them both in order to lock in correctly. The first of these blocks has an RQ in the second character, with two alphabetic characters in the other two positions.

At the same time, the second block has two more alphabetic characters in the first two positions with an RQ in the third. The RQ characters prevent the four alphabetic characters from printing out at the slave station. These four characters can be chosen by the users, but must be agreed upon beforehand by the operators at the two stations concerned. In SITOR, the commercial maritime service, these characters are derived from the ship's telex number and form a selective-calling code, or SELCAL. As used in the Amateur Radio Service (AMTOR: AMateur Teleprinting Over Radio), the four-character group selective- calling code suggested for all random communications is derived from the call letters of the station.

To accommodate any slow drift in timing between the two stations after initial contact, the slave station monitors the timing of data transitions received from the master. If these tend to drift away from the-optimum point, i.e., halfway between the adjacent sampling instants, the local clock is shifted to correct this. Thus, the slave timing follows exactly that of the master. The master uses the same technique to make sure it is sampling the signal from the slave at the optimum instants.

~~RESYNCHRONIZATION~~

Because timing drift correction is very slow in action, it is not easily disturbed by short periods of interference. But if contact is lost completely for some period of time, both stations must reestablish the correct timing by manual operator intervention - restarting the

~~261 contact as if commencing a new communication session.~~

When both stations have been receiving errors or requests for repetition for 32 blocks, then both stations will automatically return to the synchronization procedure, with the sending station retaining any unsent message in a buffer. The remarkable feature of the system is that it remembers which station was sending before the interruption, and when back in synchronization again, a change of direction is made automatically if required. The remainder of the interrupted message is then sent without gaps or errors.

~~TIMO NG CONSIDERATIONS~~

CCIR Recommendation 476, which defines STTOR in the international maritime radio service, specifies the block repetition rate at 2.222 per second and the data rate within bursts at 100 bits per second. Thus, a block of three characters takes 210 milliseconds and a control code 70 milliseconds, leaving 170 milliseconds during which neither station is transmitting.

At first glance, it might seem like a good idea to allow the biggest margin of time for delays in antenna changeover relays, and to arrange the slave station to reply 85 milliseconds after the end of the master's transmission. The effects of distance between the two stations, however, cannot be ignored. This is particularly so for intercontinental communications.

The velocity of radio waves is 186.4 miles/millisecond (300 kilometers/millisecond). As a result, the slave station will receive a delayed signal from the master, and the resultant reply will be received late at the master station by two. milliseconds for every 186.4 miles (300 kilometers) separating the two stations. Thus, to make sure that this slave reply is not obliterated by the next master transmission on long distance communications, the slave must reply as soon as possible after receiving the signals from the master.

With practical equipment, and taking into account delays through various filters in the equipment, it looks as though 12,400 miles (20,000 kilometers) is about the maximum range for ARQ to function successfully. In other words, ARQ will just about cover the world on BF, at least by short path, but rules out some satellite possibilities and moon bounce.

~~FEC - FORWARD ERROR CORRECTION - MODE B~~

In addition to ARQ, STTOR and AMTOR provide an alternate mode for situations in which the interactive or "handshake" operation is unsuitable. For example, in general broadcasts such as weather and press transmissions, traffic and message lists, emergency notices to mariners, and roundtable types of "network" operation.

~~In such cases, it is clear that synchronization and "handshaking" cannot be established~~

2-62 between one station and more than one other. Yet, the majority of services that use SITOR do have a definite need for such "broadcast" facilities. This is the role played by SITOR's FEC or Forward Error Correction mode. The form of FEC used in TOR depends on time diversity; each character is transmitted twice.

Instead of the usual bursty ARQ procedure, the sending station transmits the same character codes in a continuous stream, repeating each character twice. Each character is interleaved with the other characters, its second appearance delayed by 280 milliseconds.

The first appearance of the character is analyzed for validity, and stored in memory if valid. If the character is invalid, it is rejected and a "space" character is written into memory in its place. If tee same character is validated and accepted at its second appearance, it is output to the terminal. If the second appearance of the character is also invalid, the previously- . stored character is output to the terminal, regardless of whether it is an accepted character or a "space" character. Thus, a "space" will be printed out only when both appearances of the same character have failed validation. The human operator can usually fill in the occasional missing characters found in a word.

The main point is that the terminal or printed page will not be filled with "garbage" or mutilated data. The station equipment can remain tuned to an empty, noise-filled channel without displaying continuous streams of garbled characters on the screen or page. Until valid data is received, the terminal will display nothing.

For the station operator, FEC operation is almost identical to conventional radioteleprinter or telex service; each station transmits in turn. The error detection and correction available with FEC is not as effective as that with ARQ. However, in those cases where ARQ is not workable because of multiple station circumstances, FEC's performance characteristics greatly surpass conventional Baudot-Murray and ASCII RTTY service.

~~263 THE ORIGIN OF PACKET T ANSMISSION~~

The concepts of packet transmission and switching appeared for the first time in this country in studies of military communications networks by Paul Baran and his co-workers at the Rand Corporation in 1964. At about the same time, several European organizations were studying and planning similar types of "distributed communications" systems.

In England, Davies, Bartlett and others at the National Physical Laboratory (NPL) proposed that a store-and-forward system using short message units called "packets" would be best able to serve interactive computers, because the computers naturally generated and received short messages. The delays inherent in store-and-forward methods would be reduced by restricting the length of the packets and using high-speed lines between the switching systems. The practical outcome of the NPL work was a local packet-switched communication network which grew in a number of years to serve about 200 user terminals and give them access to about a dozen computer services.

~~The Rand Corporation work was not aimed primarily at computer communication, but led to the same kind of network design that later was proposed for packet switching.~~

The first operational packet network in the U.S. was ARPAnet, designed to interconnect university computer centers and other centers where ARPA-funded projects were in progress. (ARPA is the Advanced Research Projects Agency of the U.S. Department of Defense.) ARPAnet used TCP/IP, a set of protocols developed by a community of researchers to allow cooperating computers to share resources across a network. (The original ARPAnet has been phased out and is being replaced by a variety of networks running the same protocols loosely referred to as "The Internet".)

In Europe, parallel work by many organizations led to the realization that standards had to be derived for international operations using packet transmission and switching. Packet networks had already begun operation in the U.S. (Telenet), England, Canada (Infoswitch), Switzerland (Bernet), and the German Federal Republic.

The CCITT, (International Consultative Committee on Telephone and Telegraph), an agency of the ITU (International Telecommunications Union), charged with development of recommendations and standards, tasked Study Group VII on Data Transmission, to draft a series of recommendations that would insure the uniform development of packet transmission and switching protocols and methods. In 1976, the Plenary Session of the ITU unanimously approved CCITT Recommendation X.25, "Interface Between Data Terminal Equipment (DTE) and Data Circuit Terminating Equipment (DCE) for Terminals Operating in the Packet Mode on Public Data Networks".

Recommendation X.25 stated, "the establishment in various countries . of public data networks providing packet-switched data transmission services creates a need to produce standards to facilitate international interworking". X.25 was amended and expanded in 1980. X.25

264 refers to additional Recommendations, including X.1, X.2, X.21, X.21 bis, X.92, X.96 and other related standards. X.25 was accepted by nations signatory to the I TU treaties and was adopted by the Bell Operating Companies as Bell Standard BX.25.

~~IN'T'RODUCTION TO PACKET RADIO~~

~~WHAT IS PACKET RADIO?~~

Packet radio is a method of high-speed data transmission, in which VHF, UHF and HF radio are used to carry the packets instead of telephone circuits. Unlike conventional RTTY using either the Baudot-Murray code or ASCII, but in common with STTOR, packet radio benefits from the use of built-in error detection and error correction provisions.

PACKET transmission provides error-free transmission of any kind of data, regardless of the user's data code and speed, at the fastest possible network speeds consistent with the available bearer circuits. The packet link transmits ASCII code in the form of NRZI data. Packet systems operate most efficiently on normal voice-quality telephone and FM radio bearer circuits, at data rates far in excess of those normally used in BF radio.

When poor BF conditions are adequate for SITOR operation, packet transmissions may be subject to large numbers of error-detection retry attempts. The resulting data throughput rate could be much lower than that achieved by STTOR.

Unlike STTOR, in which networking is presently limited to several automatic but non- standard methods of multihop mailbox-to-mailbox and store-and-forward message handling, packet networks can be formed easily and automatically by virtue of the "digipeater" functions built in to some of the more advanced packet radio terminal node controllers, as well as recently developed higher level packet switches. With "digipeaters" and Level 3 switches, it is simple and cheap (and almost automatic) to establish an extensive network of packet nodes spanning a wide geographical area.

Packet radio systems consist of several types of devices - packet terminals, packet controllers, packet assembler/disassemblers, packet repeaters, packet switches and individual manually operated packet stations.

~~A packet terminal can be a simple dumb terminal, a microcomputer with a smart terminal emulation program, a host computer, or a device somewhere between these categories.~~

The terminal or computer is connected to some form of packet controller and a. radio transmitter/receiver system. These packet controllers are also known as TNC or Terminal Node Controller, and PAD or Packet Assembler/Disassembler).

~~A packet repeater provides radio paths between terminals and stations, or between pairs of terminals. In the general case, the packet repeater receives, stores, inspects, and retransmits packets.~~

~~A packet switch provides automatic routing of virtual circuits on one or more radio channels, and may use real-time switching or store-and-forward operation.~~

~~A packet station can perform routing, flow control, logging. accounting, and other functions for the whole system.~~

~~THE BEGMING OF PACKET RADIO~~

Much of today's packet radio technology has sprung from developments of the ground radio system called ALOHA. We'll look as this system first, then proceed to a discussion of the present packet radio systems.

The ALOHA system was a UHF packet broadcast system created for practical reasons, including the poor quality of local telephone lines, by a team at the University of Hawaii. It was first operational in 1970, covering the Hawaiian Islands centered on the island of Oahu. Inexpensive access was afforded to central time-sharing computer systems for several hundred terminal users. In the first instance, communication was limited to a group of terminals in the Honolulu district within direct radio range of the central station. User-to- user communication was also available.

ALOHANET was designed to provide cheap and easy access to central computing facilities for a large number of terminal users. A summary of the ALOHA project can be found in Binder et al. User-to-computer communication was by means of a 100-kHz random access channel at 407.350 MHz; the broadcast return channel, computer-to-user, was also 100-kHz bandwidth at 413.475 MHz. Direct user-to-user communication was not provided; user-to- user communication was possible by transferring data to the central switch and forwarding it to the destination user. Until the addition of packet repeaters, the system was logically equivalent to a star-connected network.

The central communications processor, an HP 2100 minicomputer, at Honolulu on Oahu, received packets from users and was responsible for sending packets to them. The packet transmission data rate was 9600 bits per second. Packets consisted of a header (32 bits), a header parity check field (16 bits), and up to 80 bytes of data, followed by a data parity check field (16 bits). Maximum size packets were therefore 704 bits long and took about 73 milliseconds to transmit; propagation time was negligible in comparison.

Control of the broadcast channel from the central computer to the users presented no problem, because only one transmitter used the channel. The packet headers contained user addresses which enabled individual receivers to identify the traffic intended for them. The random access user-to-computer channel could have been apportioned to individual users by a fixed allocation scheme, such as frequency division multiplexing or time division multiplexing. However, the nature of terminal traffic was almost always bursty and a fixed allocation would hardly have made the best use of the communication medium, hence the choice of the random access scheme.

This scheme, known as pure ALOHA, allowed a packet terminal to transmit packets at times completely independent of packet transmissions from other terminals. A natural consequence of this independence was that packets from different sources could be transmitted at the same time and therefore collide or overlap as they arrived at the central station. An overlap that ' affected only the smallest fraction of transmission time had the same effect as an overlap of complete packets; both packets were irretrievably corrupted.

Packets subjected to such overlap were rejected by the central station and the collision was made know to the respective transmitting terminals by the absence of the acknowledgement signal which would otherwise have been sent by the central station to the packet terminals.

Packets refused by the central station were retransmitted by the packet terminals after a timeout period. Immediate retransmission of packets by these terminals, or indeed, retransmission after a fixed, uniform time interval, would have just resulted in a second overlap; therefore, retransmission took place at each terminal after a random delay.

Clearly the number of collisions was a result of traffic intensity; the greater the traffic, the greater the probability of overlaps. It was also essential that acknowledgement packets be sent with highest nonpreemptive priority from the central station to the packet terminals; otherwise unwanted retransmission might have occurred.

Error control on the broadcast channel (central station to packet terminals) presented difficulties. Ideally, this should have been on the same positive acknowledgement basis as the error control in the other direction on the random access channel. However, acknowledgement packets destined to the central station would have had to contend for the random access channel the same way as data packets.

Because of this contention, at full channel loading each random access packet had to be retransmitted an average of 1.7 times. Thus, each data packet or acknowledgement packet had to be sent an average of 2.7 times before it was received successfully.

In an error-free situation, to ensure that the acknowledgement is successfully transmitted by the packet terminal, the central station had to send the data packet an average 2.7 times, although the packet might have arrived correctly the first time.

Where errors occur, the multiple transmission from the central station will be essential if an acknowledgement system is to operate correctly. This is very wasteful of bandwidth and can be avoided by not using acknowledgements in this channel, relying on low error.rates, and a system of reporting errors to the user who may decide to repeat a transaction. Where this is

~~267 not acceptable for specific users, a system of positive acknowledgement may be introduced on a selective basis.~~

We have already observed that immediate or constant-delay retransmission of collided packets would result in inevitable further collisions. Therefore, retransmission must take place only after a time interval to be chosen at random by the packet terminals.

Because the maximum propagation delay in a ground radio packet broadcast system is small compared with the packet transmission,time (say, one/tenth or less), better channel utilization is obtainable if the packet terminal wishing to transmit a packet first listens to the channel before transmitting.

If the terminal wishing to transmit detects that the channel carrier is active, it then waits until the channel becomes quiet, testing for this at intervals. This is the principle commonly used for aircraft voice communication. However strictly the "listen before talk" rule is applied in packet radio, there may be collisions because of the time it takes for the packet from one terminal to reach another terminal. If the unguarded interval is very small compared with the time taken to transmit the packet, the probability of collision need not be high; but if many stations are waiting to jump in at the end of anther's transmission, collision could be more frequent than necessary.

In the packet radio context, the mode of operation in which packet terminals listen to the broadcast channel has come to be known as Carrier Sense Multiple Access (CSMA). The principle was first suggested by Wax of the University of Hawaii in an internal memorandum of 1971.

We have seen how packet radio was first developed in the ALOHA sense, with packet radio terminals all within direct radio communication with one central packet radio station. pure and slotted ALOHA channel multiplexing techniques are primarily applicable to this sort of network; two channel frequencies are used, one to the central station and one from it.

These techniques can be applied to a more general packet radio in which individual packet terminals need not be within radio range of each other; packet radio repeaters may be used to collect traffic from packet radio terminals and to hand these on, either to other repeaters or to a central packet radio station. This network can also be used as a general communication network rather than as a star network with central computing facility. Traffic can pass from terminal to terminal, and may or may not need to travel via a central packet radio station. One common channel frequency may be used for all traffic, whatever its direction.

In the most basic form of CSMA, the packet terminal with a packet to transmit, commonly called a "ready" terminal, senses the channel and, if it finds it idle, immediately transmits its packet. If the channel is not idle, the packet terminal schedules a further attempt at transmission at a random time with a probability distribution obtained by sampling a

268 specified retransmission delay distribution. When the time for another attempt arrives, the packet terminal repeats the process. --In this simple form of CSMA the broadcast channel is not time slotted and transmissions may start at any time. This mode of CSMA operation is termed "non-persistent" by Kleinrock and Tobagi; because of the ability to begin transmission without reference to a synchronous time frame, it is further categorized as "unslotted".

~~THE ABCs OF PACKET RADIO~~

Where specific references are made to packet radio procedures, protocols and practices, to avoid confusion, these discussions of packet radio techniques will be based on the widely- accepted AX.25 packet protocols. We will also discuss the use of the Internet TCP/IP protocol suite.

The AX.25 packet specification was adapted from MIT Recommendation X.25, the international standard for packet protocols developed in 1977 by the NU at Geneva, Switzerland, for use in the commercial landline packet networks of the world.

Packet radio systems operate on a common communication channel. Each packet can be received at a number of other packet terminals. The packet header is inspected at each terminal to determine whether or not the packet is destined for a particular terminal. Packets addressed to other terminals are ignored; those addressed to the terminal are passed on to it. All terminals listen to the same channel; therefore, the reception of an acknowledgement packet by the original transmitting terminal indicates correct reception at the desired destination terminal.

Packet radio is vulnerable to clashes in transmission which can occur when two or more transmitting terminals overlap. In packet radio, the transmitting station's controller can listen to the radio channel before transmitting and then transmit only when the radio channel is clear. This procedure known as CSMA (Carrier Sense Multiple Access) tends to reduce collisions but does not prevent clashes completely because simultaneous or almost simultaneous transmissions can still take place.

~~In all the systems, solving the contention for use of the transmission medium is preeminently important. A great deal of thought has gone into designing systems to perform this function efficiently.~~

~~AMATEUR RADIO STANDARD AXIS~~

Amateur standard AX.25 broadly follows the provisions of Recommendation X.25, complies with link access procedures and almost all other protocol methods. The area in which AX.25 departs significantly from X.25 is in the method of addressing the destination station

~~269 and the originating station. X.25 uses a standardized international numbering system (Recommendation X.121) suitable for the common carrier networks.~~

AX-25 modified the addressing scheme to use the amateur radio operator's call letters. The originating station can direct its packets to a destination station and route the packets via up to eight packet repeaters (digipeaters). This provides an integrated, rational method of establishing network routing, and permits automatic linked operation of individual packet controllers as digipeaters.

~~PACKET FUNDAMENTALS~~

~~WHAT IS A PACKET FRAME~~

A packet frame is like an envelope into which data are placed. In addition to the data typed at the user's keyboard, the envelope contains synchronizing and addressing information, as well as error-detection control information.

A packet frame can also be described as a sequence of data bytes or "octets", (binary groups), formatted in a special manner, strung together and transmitted as a single piece of information from one end of a network to another, from one user to another user. If the user's message or data file is too long to fit into a single packet frame, the message or file can be cut up automatically by the packet controller, divided into several packets and transmitted sequentially.

Each user must be equipped with a PAD (Packet Assembler-Disassembler) or TNC (Terminal Node Controller), a device that assembles his terminal's output data into packets for outbound transmission, and disassembles the inbound data from packets into data suitable for display on the user's terminal.

~~TSE HDLC FRAME~~

As a result of work done at IBM by J, R. Kersey and others in the early 1970s, a protocol was proposed and later adopted as an international standard by the International Standards Organization (ISO) in 1979. That standard is called the FEgh Level Data Link Control (HDLC) procedure.

In HDLC, a message synchronization indicator (the flag) is generated by hardware circuits. Other hardware circuits prevent any data being transmitted from having the same pattern as the flag. The flag then becomes sort of an out-of-band framing signal, like the start and stop bits in the asynchronous teleprinter protocols. Because the data being transmitted are examined on a bit-by-bit basis to screen out any possible aliases of the flag, HDLC and

~~270 similar protocols such as IBM's Synchronous Data Link Control (SDLC), are referred to as "bit-oriented protocols" (BOPS). -~~

~~In HDLC, all information is carried by frames which may be of three types:~~

~~Information frames (I-frames): I-frames perform information transfer, and independently carry message acknowledgements, and poll or final bits.~~

~~Supervisory control frames (S-frames): S-frames perform link supervisory control such as message acknowledgements, retransmit requests, and requests for temporary holds on I-frame transmissions.~~

Unnumbered command/response frames (U-frames): U-frames provide a flexible format for additional link control data by omitting the frame sequence numbers and thus providing a place for additional command and response functions.

~~WHAT'S IN A PACKET FRAME?~~

A packet frame consists of different data groups called "fields", in a fixed, predetermined sequence. Each data group or field represents a different kind of information and is used for a different purpose. A typical packet frame will include the following fields:

~~THE FLAG~~

The flag contains a single "byte", or digital character, made up of a special sequence of bits (binary digits) - 01111110. This exact sequence of bits, a unique combination that will not be duplicated in normal text and data files, is always used by every packet controller.

The flag is the first field in the packet frame and provides synchronizing information to the distant system, in effect telling the other system where the packet frame begins and ends, and providing a reference "clock" for synchronization of the two or more systems in the packet network. The flag is also transmitted a second time, as the very last field in the frame, and identifies the end of the packet frame.

~~THE ADDRESS FIELD~~

~~The X.25 address field contains a single octet or eight-bit byte, and identifies both the source of the frame and its destination. CCITT Recommendation X.121 provides the addressing~~

~~271 format under the International Numbering Plan used in the telephone packet networks.~~

The AX.25 address field, as modified in the Amateur Radio Service, contains from 14 to 70 bytes, and carries addressing information for the destination station and the originating station. In addition, the address field contains command/response information and facilities for the unique Level 2 repeater operation.

Rather than the X.121 numbering plan, the AX.25 address uses the identifier or call sign of the station to whom the packet is addressed, the call sign of the originating station, and call signs of up to eight "digipeaters" (packet repeaters) that can be designated by the user to create automatic routing of the packets along a temporary network.

~~THE CONTROL FIELD~~

The control field contains one octet, and identifies the type of frame being passed and control several attributes of the connection. This field contains information about the nature packet, of the the type of frame, acknowledgement of a good packet, rejection of a bad packet, the number of the packet in a sequence of transmitted packets, and other protocol- oriented functions needed to control flow of packets from station to station and along a packet network.

~~THE INFORMATION FIELD~~

The information field usually contains user data, the information you originate by typing at your terminal's keyboard. Recommendation X.25 and Amateur Radio AX.25 permit field lengths up to 256 octets. The information field is almost always an integral multiple of one octet or byte in length.

Present operating practices generally call for an information field length of up to 128 bytes or message characters. If the user types less than 128 characters before a "return" or "enter" keystroke, the information field will contain those characters typed up to the "return" or "enter". The "return" or "enter" keystroke is usually the character that commands the packet controller to format the packet and frame, send the frame to the telephone line, or in the packet radio application, key the transmitter, and send the packet.

If the user types more than 128 characters without a "return", the system will end the information field, format the-packet frame and transmit the frame to the radio or the network. The remaining characters typed by the user will be automatically carried over and sent in a subsequent packet frame. In some types of packet frames used in link control, the information field may not exist, or.may be empty.

~~'272 THE FRAME CHECK SEQUENCE~~

The frame check sequence field contains a 16-bit digital word developed automatically by the packet controller for transmission to the distant station. The sending packet controller's microprocessor performs a mathematical operation, the results of which are placed in the frame check sequence and transmitted to the distant end. The receiving packet controller examines the received packet, calculates the frame check sequence from the incoming packet; and compares the result with the value sent by the originating packet controller. If the two values fail to match, the receiving packet controller discards the received packet and requests", repetition of the packet from the originating station.

~~THE FLAG ... AGAIN~~

~~This second flag is a repetition of the flag sent at the start of the frame, and contains the same synchronization byte, "01111110". The flag closes the packet frame.~~

~~THE TELEPHONE ANALOG OF PACKET RADIO~~

Perhaps the easiest way to picture the manner in which packet radio works is a hypothetical comparison to an ordinary telephone call. The following example illustrates the concepts of communications protocols, network access, and control procedures.

~~The term "communications protocol" defines a set of procedures by which communication is accomplished within standard restrictions.~~

In our packet radio case, we have an operator's terminal, a transmitter and receiver, a packet controller and a modem. These pieces of equipment form the basic package at each end of the communications link.

The communications link is divided into two different functions: the "circuit connection" and the "link connection". A "circuit connection" is simply an electrical path created between two or more points that wish to communicate. The connection can be metallic (wires or cables), and/or radio. The fact that we create an "electrical" path or connection does not necessarily mean that communication can occur. A "link connection" is created by procedures that prepare the originating station to transmit information, and the receiving station to receive that information.

An ordinary telephone call has three separate parts: the call setup, during which the connection is requested and established; the transfer of information, during which the calling party and the called party hold their conversation; call termination, during which the parties to the call disconnect from each other.

~~2 f3 In his excellent book, "Telecommunication System Engineering", Roger Freeman sets up a clever comparison between a routine telephone call and an analogous data communication:~~

1. CALLING PARTY (C): "Hello" 2. ANSWERING PARTY (A): "Hello" 3. C: "Good morning. may I speak to Paul Jones?" 4. A: "Just a moment, I'll see if he is in; who may I say is calling?" 5. C: "John Doe" 6. A: "Just a moment, please" , 7. A: "Good morning, John" 8. C: "Good morning, Paul. If you can spare a minute, perhaps on...." we can settle the matter 9...... (talking) 10...... (talking) 11. C: "Then that's settled, and I'll confirm by letter." 12. A: "Thank you, John; I appreciate the call, and we will be talking again soon." 13. C: "Goodbye." 14. A: "So long for now" (conversation terminated) (Both parties hang up - connection terminated)

Freeman notes that steps 1 through 8 initiate the conversation; the calling and called parties are identified. On data and packet networks, this is called "handshaking". - Step 8 is the start of text, and step 11 is the end of text. Step 12 may be considered the positiw acknowledgement (ACK), and steps 13 through 15 are analogous to the end of message (EOM).

Try to picture this type of formalized exchange taking place as a data transmission on a VHF band, during which you are typing at your keyboard, while your communication session partner types at his or hers. A pair of packet controllers set up and control transmission and reception, validate and eventually terminate the connection (link) over which you held your communication session at 1200 bauds.

~~Here is a general idea of what a packet radio communication session might appear on my video display if you were looking over my shoulder as I let my fingers do the talking.~~

~~Assuming that my packet controller is set with my call sign, my monitor is enabled and my radio equipment is up and operating, I make the first call. I want to call AABA, so I type:~~

~~CONNECT AABA < cr/lf >~~

My transmitter is keyed on and off by my packet controller as it transmits a series of connect-request packets, until either AABA's packet controller responds, or my system exceeds a programmable number of automatic connect request retrys. If AABA's packet

274 controller and associated radio gear are active and set to accept a connection request, and I have a reasonable signal path to his .location, after my radio has keyed on and off a couple of times and sent a couple of connect request packets, I will probably see the following on my screen:

*** CONNECTED TO AABA

~~Now, I can sit and type my message to AABA, or just chat, or just see how the system is working.~~

~~SITOR VERSUS PACKET RADIO - WHICH IS BEST?~~

Debate has arisen during the last several years regarding the differences between SUOR and packet radio as "competing" forms of digital communications. There is really only one important characteristic common to both of these RTTY modes: both SI TOR and packet radio have built-in systems of error detection and error correction. Beyond that, these modes are very different; each mode is best suited to the specific applications for which it was designed. SITOR and packet radio are NOT competing for your attention!

~~HANDLING MESSAGE + + C? -CHOOSE Sam~~

~~SUOR is designed to get a text message (such as conventional "radiogram" traffic) through in the poor propagation and interference conditions common to BF operation.~~

Based on its origins in telex, SUOR is limited by regulation to a signalling speed of 100 bauds, resulting in a throughput speed of 50 bauds or 66 WPM: At the present time, no higher-speed SUOR protocols exist, although experimental work is being done with SUOR at higher speeds.

The Netherlands Post and Telegraph Administration and the Dutch electronics firm Philips, developed SUOR specifically to bring high-seas vessels into the existing international telex network via HF radio. The CCI1T international standards of performance for the world's telex networks specify that the error rates in the telex network may not exceed three character errors in 100,000 characters.

The seven-bit SUOR code was derived from the 5-bit Baudot-Murray code (known as Can International Telegraph Alphabet Number 2) and is easily translated to and from that code. The use of ITA #2 is mandatory in the world-wide telex network, although the code contains only 32 character combinations. Two of these characters are used as control characters, shifting between letters and figures to allow upper-case letters, numerals, and limited typewriter-style punctuation.

275 The two additional bits in the STTOR code theoretically permit 128 character combinations. However, in the STTOR system, this is reduced to 35 combinations, (admitting the same characters used in the Baudot-Murray code), as a result of the elimination of all possible combinations except those character data strings that can follow a constant-pattern mark-to- space ratio of four marks to three spaces (4/3). The strict use of this 4/3 mark/space ratio provides significant error detection and error correction when messages are sent in blocks of three characters.

SITOR does not use any form of parity, CRC or checksum to detect errors. Thus there is not much overhead ("excess baggage" - complex control characters) in information being transmitted. There is some statistical evidence showing that invalid characters may be falsely validated approximately once in every 20,000 characters - an acceptably low probability.

In ARQ, when the Information Receiving Station (IRS) receives a block with three valid characters, it sends a control signal back to the Information Sending Station (ISS). The returned control character commands the ISS either to send the next block of new text, or to repeat the previous block of three characters. During worst-case path conditions, the link between the two stations may fail entirely.

The system will stop when only invalid data is being passed, or when requests for repetition are continuous. When this occurs, the SI TOR protocols require that the station which initiated the communications link (the "master") automatically revert to a resynchronization routine until the path is once again adequate for passing error-free messages.

Thus, under extremely poor conditions on BF, the traffic will be moved, even between the dots and dashes, of interfering CW signals. The link throughput rate (the amount of valid information actually transferred) will slow down, perhaps drastically (maybe only five WPM!). But, in many cases, slow error-free throughput is better than no transfer of data at all. Any minimum path will allow traffic to be moved!

SITOR, as accurate as it may be, cannot be used normally to transfer computer program or binary files. The SITOR code set does not include many of the characters required for program listings. For example, source code in BASIC, Pascal and C languages require characters like the "greater than" and "lesser than" (> <) signs, "asterisks" (*), "equals" signs (=), left and right "square brackets" (f ]), left and right "curly brackets" ({ )) and other characters contained in the ASCII and EBCDIC code sets, but absent from the Baudot- Murray and SPTOR codes.

~~276 DATA TRANSFER OR HEAVY MESSAGE TRAFFIC UNDER GOOD PROPAGATION CONDITIONS - PACKET RADIO IS THE BEST CHOICE!~~

~~Packet radio, unlike SITOR, can move any type of information in any character code in a network environment. The packet protocol is transparent to the code and speed used by the sending station.~~

The packet equipment (TNC, PAD, hardware and software) converts the sending station's data into the form the network requires, and at the distant end, reconverts the data back to the form required by the receiving station. Your terminal can literally operate in Baudot- Murray, ASCII, EBCDIC - the network doesn't care. Packet radio is the ideal means of transferring computer program and binary files, with or without modem file transfer protocols.

Error detection in packet protocols uses a special mathematical expression, which, in effect, calculates a numerical value representing each transmitted data character; the values of all the characters in the sync (flag) field, the control and address fields, and the entire information field (the data or information typed by the user). The derived value is placed in the frame check sequence field and the packet is transmitted to the network.

The receiving station's system performs the same polynomial calculation of the received packet frame containing the same flag, control, address, and user data fields, derives the frame-check value, and then compares the values of the received data to the value contained in the received frame check sequence field. If these values agree, the receiving station sends an "acknowledge" packet backwards on the network and the sending station continues with the next packet.

If the frame-check sequence value calculated by the receiving station does not agree with the value in the received frame-check sequence, the receiving station throws the packet away and sits there listening. The transmitting station counts the time since the transmission of the last packet, and after the system decides that enough time has elapsed and it has not received the "ACK" packet, the sending station retransmits the same packet again and once more waits for the "ACK". This automatic "retry" operation will continue for up to a predetermined number of retries, after which the sending system decides that the link is failed, issues a "disconnect" command to the distant end, shuts down and waits for a new operator command.

Packet systems are designed to operate over a standard voice-grade circuit, that is, a circuit equivalent to a voice-quality telephone line. This is the quality of signal generally available on a reasonable VHF/UHF FM link, or as achieved on the higher BF channels during good propagation conditions.

277 Packet radio was not meant to be used in environments with noise, selective fading, sideband phase distortion, heterodynes from- interfering signals, electrical static, over-the-horizon pulse radars, etc. While the Amateur Radio Service is making significant use of packet radio at 300 bauds on HF, there is growing evidence that shows packet to be less effective than SITOR for passing text message traffic under HF's present severe conditions.

The packet network protocol sets the maximum number of unacknowledged packets that can "fly" on the network at the same time. The AX.25 protocol permits a maximum of seven unacknowledged packets (the "window"), and usually up to 15 retries.of an unacknowledged packet before issuing the "disconnect" command. (It is possible to set a link for an infinite number of retries although this is not a generally recommended practice.)

Because the receiving station must validate perhaps up to several hundred data characters in the block (packet frame), the pure mathematical probabilities of rejecting the packet frame are several orders of magnitude higher than in SITTOR. The chances of having a wrong data character inside a frame have been estimated at 1 in 65,000 if the packet frame check sequence is validated.

Rather than receipt of a wrong character, there should be more concern with the loss of an entire packet, which frequently happens. If the link times out and is disconnected due to excessive retries, and many new "reconnects" are needed, the chances are very high that data will be lost. The unacknowledged packets simply vaporize into the great beyond and are gone forever.

The real questions are: "how many times will the packet link time out, be disconnected and a new connection required before all the data is transferred?", and "what is the net throughput rate?". How long does it really take to get error-free information transferred?

~~IS SITOR TOO SLOW?~~

SITOR has been accused of being "too slow for transferring a lot of information". In fact, under the conditions typically suitable for the use of 45-baud (60 WPM) Baudot-Murray MY, SITOR running at the mandatory 100 bauds (throughput rate of 66 WPM) is about 10 percent faster! Under the poor conditions which would result in substantial errors in Baudot-Murray and ASCII RTTY, SITOR may slow down because of block repeats. Few, if any, errors will occur! Under the same conditions on HF, packet radio would probably not get the shortest information packet through (that would be at least 17 bytes, all overhead without user data).

~~278 TCP/IP~~

TCP/IP is a set of packet protocols developed to allow cooperating computers to share resources across a network. It was developed by a community of researchers centered around the ARPAnet, the best-known TCP/IP network. However as of June, 1987, at least 130 different vendors support TCP/IP, and thousands of networks of all kinds use it. The original ARPAnet has been phased out and is being replaced by a variety of networks running the same protocols loosely referred to as "The Internet".

~~TCP/IP BASICS~~

~~The most accurate name for this set of protocols is the "Internet protocol suite'. TCP and IP are two of the protocols in this suite.~~

~~Because TCP and IP are the best known of the protocols, it has become common to use the term TCP/IP or IP/TCP to refer to the whole family.~~

The Internet is a collection of networks, including the Arpanet, NSFnet, and regional networks such as NYsernet, local networks at a number of University and research institutions, and a number of military networks.

The term "Internet" applies to this entire set of networks. A subset of them, managed by the Department of Defense, is referred to as "DDN" (Defense Data Network). This includes some research-oriented networks such as the Arpanet, as well as more strictly military ones.

All these networks are connected to each other. Users can send messages from any of them to any other, except where there are security or other policy restrictions on access. Officially speaking, the Internet protocol documents are simply standards adopted by the Internet community for its own use.

Recently, the Department of Defense issued a MIMPEC definition of TCP/IP, intended as a more formal definition, appropriate for use in purchasing specifications. However, most of the TCP/IP community continues to use the Internet standards. The MIISPEC version is intended to be consistent with it.

TCP/IP is a family of protocols. Some provide "low-level" functions needed for many applications. These include IP, TCP, and UDP. Others are protocols for doing specific tasks, e.g., transferring files between computers, sending mail, or finding out who is logged in on another computer.

~~Initially TCP/1P was used mostly between minicomputers or mainframes. These machines had their own disks, and generally were self-contained.~~

~~279 The most important "traditional" TCP/IP services are:~~

~~FILE TRANSFER - FTP~~

The File Transfer Protocol (FTP) allows a user on any computer to get files from another computer, or to send files to another computer. Security is handled by requiring the user to specify a user name and password for the other computer. Provisions are made for handling file transfer between machines with different character set, end of line conventions, etc. FTP is a utility that you run any time you want to access a file on another system. You use it to copy the file to your own system. You then work with the local copy.

~~REMOTE LOGIN~~

The Network Terminal Protocol (IM NET) allows a user to log in on any other computer on the network. You start a remote session by specifying a computer to connect to. From that time until you finish the session, anything you type is sent to the other computer. Note that you are really still talking to your own computer. But the Telnet program effectively makes your computer invisible while it is running.

Every character you type is sent directly to the other system. Generally, the connection to the remote computer behaves much like a telephone dialup connection. The remote system will ask you to log in and give a password, in whatever manner it would normally ask a user who had just dialed it up. When you log off of the other computer, the Telnet program finishes. Microcomputer implementations of Telnet generally include a terminal emulator for a common type of terminal. By the way, the Telnet protocol should not be confused with Telenet, a vendor of commercial network services.)

~~COMPUTER MAIL~~

This allows you to send messages to users on other computers. Originally, people tended to use only one or two specific computers. They would maintain "mail files" on those machines. The computer mail system is simply a way for you to add a message to another user's mail file.

There are some problems with this in an environment where microcomputers are used. The most serious is that a micro is not well suited to receive computer mail. When you send mail, the mail software expects to be able to open a connection to the addressee's computer, in order to send the mail. If this is a microcomputer, it may be turned off, or it may be running an application other than the mail system. For this reason, mail is normally handled by a larger system, where it is practical to have a mail server running all the time.

~~230 Microcomputer mail software then becomes a user interface that retrieves mail from the mail server.~~

These services should be present in any implementation of TCP/IP, except that micro- oriented implementation's may not support computer mail. These traditional applications still play a very important role in TCP/IP-based networks. However more recently, the way in which networks are used has been changing. The older model of a number of large, self- sufficient computers is beginning to change. Now many installations have several kinds of computers, including microcomputers, workstations, minicomputers, and mainframes. These computers are likely to be configured to perform specialized tasks. Although people are still likely to work with one specific computer, that computer will call on other systems on the net for specialized services. This has led to the "server/client" model of network services.

~~A server is a system that provides a specific service for the rest of the network.~~

~~A client is another system that uses that service. (Note that the server and client need not be on different computers. They could be different programs running on the same computer.)~~

~~Here are the kinds of servers typically present in a modern computer setup. Note that these computer services can all be provided within the framework of TCP/IP.~~

~~NETWORK FILE SYSTEMS~~

~~Systems can access files on another computer in a somewhat more closely integrated fashion than FIT.~~

A network file system provides the illusion that disks or other devices from one system are directly connected to other systems. There is no need to use a special network utility to access a file on another system. Your computer simply thinks it has some extra disk drives. These extra "virtual" drives refer to the other system's disks.

This capabilityis useful for several different purposes. It lets you put large disks on a few computers, but still give others access to the disk space. Aside from the obvious economic benefits, this allows people working on several computers to share common files. It makes system maintenance and backup easier, because you don't have to worry about updating and backing up copies on lots of different machines.

~~A number of vendors now offer high-performance diskless computers. These computers have no disk drives at all. They are entirely dependent upon disks attached to common "file servers".~~

~~281 REMOTE PRINTING~~

~~You can access printers on other computers as if they were directly attached to yours. ('The most commonly used protocol is the remote lineprinter protocol from Berkeley Unix.)~~

~~REMOTE EXECUTION~~

~~You can run a specific program on a different computer. This is useful when you can do most of your work on a small computer, but a few tasks require the resources of a larger system.~~

There are a number of different kinds of remote execution. Some operate on a command-by- command basis. That is, you request that a specific command or set of commands should run on some specific computer. (More sophisticated versions will choose a system that happens to be free.) However, there are also "remote procedure call" systems that allow a program to call a subroutine that will run on another computer. (There are many protocols of this sort.)

~~NAME SERVERS~~

In large installations, there are a number of different collections of names that have to be managed. This includes users and their passwords, names and network addresses for computers, and accounts. It becomes very tedious to keep this data up to date on all of the computers. Thus the databases are kept on z small number of systems. Other systems access the data over the network. This is now a required part of any TCP/IP implementation.

~~1 R2 MAL SERVERS~~

~~Many installations no longer connect terminals directly to computers. Instead they connect them to terminal servers.~~

~~A terminal server is simply a small computer that only knows how to run Telnet (or some other protocol to do remote login).~~

your If terminal is connected to one of these, you simply type the name of a computer, and you are connected to it. Generally it is possible to have active connections to more than one computer at the same time. The terminal server will have provisions to switch between connections rapidly, and to notify you when output is waiting for another connection. (Terminal servers use the Telnet protocol, already mentioned. However any real terminal

~~282 server will also have to support name service and a number of other protocols.)~~

~~NETWORK-ORIENTED VVEgWW SYSTEMS~~

Until recently, high-performance graphics programs had to be executed on a computer that had a bit-mapped graphics screen directly attached to it. Network window systems allow a program to use a display on a different computer.

~~Full-scale network window systems provide an interface that lets you distribute jobs to the systems that are best suited to handle them, but still give you a single graphically-based user interface.~~

Note that some of the protocols described above were designed by Berkeley, Sun, or other organizations. Thus they are not officially part of the Internet protocol suite. However they are implemented using TCP/IP, just as normal TCP/IP application protocols are.

Because the protocol definitions are not considered proprietary, and because commercially- supported implementations are widely available, it is reasonable to think of these protocols as being effectively part of the Internet suite.

Note that the list above is only a sample of the sort of services available through TCP/IP. However it does contain the majority of the "major" applications. The other commonly-used protocols tend to be specialized facilities for getting information of various kinds, such as who is logged in, the time of day, etc.

~~TCP/IP PROTOCOLS - GENERAL DESCRIPTION~~

~~TCP/IP is a layered set of protocols. In order to understand what this means, it is useful to look at an example.~~

A typical situation is sending mail. First, there is a protocol for mail. This defines a set of commands which one machine sends to another, e.g., commands to specify who the sender of the message is, who it is being sent to, and then the text of the message. However this protocol assumes that there is a way to communicate reliably between the two computers.

Mail, like other application protocols, simply defines a set of commands and messages to be sent. It is designed to be used together with TCP and IP. TCP is responsible for making sure that the commands get through to the other end. It keeps track of what is sent, and retransmits anything that did not get through.

~~If any message is too large for one datagram, e.g., the text of the mail, TCP will split it up into several datagrams, and make sure that they all arrive correctly. Because these functions~~

283 are needed for many applications, they are put together into a separate protocol, rather than being part of the specifications for sending mail. You can think of TCP as a library of routines that applications can use when they need reliable network communications with another computer.

Similarly, TCP calls on the services of IP. Although the services that TCP supplies are needed by many applications, there are still some kinds of applications that don't need them. However there are some services that every application needs. So these services are put together into IP.

~~As with TCP, you can think of IP as a library of routines that TCP calls on, but which is also available to applications that don't use TCP.~~

This strategy of building several levels of protocol is called "layering". We think of the applications programs such as mail, TCP, and IP, as being separate "layers", each of which calls on the services of the layer below it.

~~Generally, TCP/IP applications use 4 layers:~~

~~an application protocol such as mail~~

~~a protocol such as TCP that provides services need by many applications~~

~~IP, which provides the basic service of getting datagrams to their destination~~

~~the protocols needed to manage a specific physical medium, such as Ethernet or a point to point line.~~

~~TCP/IP is based on the "catenet model".~~

This model assumes that there are a large number of independent networks connected together by gateways. The user should be able to access computers or other resources on any of these networks. Datagrams will often pass through a dozen different networks before getting to their final destination.

The routing needed to accomplish this should be completely invisible to the user. As far as the user is concerned, all he needs to know in order to access another system is an "Internet address". This is an address- that looks like 128.6.4.194. It is actually a 32-bit number. However it is normally written as 4 decimal numbers, each representing 8 bits of the address. (The term "octet" is used by Internet documentation for such eight-bit chunks. The term "byte" is not used, because TCP/IP is supported by some computers that have byte sizes other than eight bits.)

~~284 Generally the structure of the address gives you some information about how to get to the system.~~

We normally refer to systems by name, rather than by Internet address. When we specify a name, the network software looks it up in a database, and comes up with the corresponding Internet address. Most of the network software deals strictly in terms of the address.

~~THE TCP/IP DATAGRAM~~

TCP/IP built is on "connectionless" technology. Information is transferred as a sequence of "datagrams". A datagram is a collection of data that is sent as a single message. Each of these datagrams is sent through the network individually. There are provisions to open connections (i.e. to start a conversation that will continue for some time). However at some level, information from those connections is broken up into datagrams, and those datagrams are treated by the network as completely separate.

For example, suppose you want to transfer a 15000 octet file. Most networks can't handle a 15000 octet datagram. The protocols will break this up into something like 30 500-octet datagrams. Each of these datagrams will be sent to the other end. At that point, they will be put back together into the 15000-octet file. However while those datagrams are in transit, the network doesn't know that there is any connection between them. It is possible that datagram 14 will actually arrive before datagram 13. It is also possible that somewhere in the network, an error will occur, and some datagram won't get through at all. In that case, that datagram has to be sent again.

~~Note that the terms "datagram" and "packet" seem to be nearly interchangeable. Technically, datagram is the right word to use when describing TCP/IP.~~

~~A datagram is a unit of data, which is what the protocols deal with. A packet is a physical thing, appearing on an Ethernet or some wire.~~

~~In most cases a packet simply contains a datagram, so there is very little difference. However they can differ.~~

When TCP/IP is used on top of X.25, the X.25 interface breaks the datagrams up into 128_ byte packets. This is invisible to ]EP, because the packets are put back together into a single datagram at the other end before being processed by TCP/IP. In this case, one IP datagram would be carried by several packets. However with most media, there are efficiency advantages to sending one datagram per packet, and so the distinction tends to vanish.

~~Two separate protocols are involved in handling TCP/IP datagrams.~~

~~TCP (the "transmission control protocol") is responsible for breaking up the message into~~

~~285 datagrams, reassembling them at the other end, resending anything that gets lost, and putting things back in the right order.~~

~~IP (the "internet protocol") is responsible for routing individual datagrams.~~

~~It may seem like TCP is doing all the work; in small networks that is true. However in the Internet, simply getting a datagram to its destination can be a complex job.~~

A connection may require the datagram lo go through several Rutgers networks, a serial line to John von Neuman Supercomputer Center, a couple of Ethernets there, a series of 56-KB phone lines to another NSFnet site, and Ethernets on another campus. Keeping track of the routes to all of the destinations and handling incompatibilities among different transport media turns out to be a complex job.

~~Note that the interface between TCP and IP is fairly simple. TCP simply hands IP a datagram with a destination. IP doesn't know how this datagram relates to any datagram before it or after it.~~

~~It isn't enough to get a datagram to the right destination. TCP has to know which connection this datagram is part of.~~

~~This task is referred to as "demultiplexing." There are several levels of demultiplexing going on in TCP/IP. The information needed to do the demultiplexing is contained in a series of "headers".~~

A header is a few extra octets tacked onto the beginning of a datagram by some protocol in order to keep track of it. It's like putting a letter into an envelope and putting an address on the outside of the envelope. Except with modern networks it happens several times. It's as though you put the letter into a little envelope, someone else puts that into a bigger envelope, the departmental mail center puts that envelope into a still bigger one, etc.

~~286 BIBLIOGRAPHY~~

~~International Standards Organization, "REFERENCE MODEL OF OPEN SYSTEMS ARCHITECTURE," Document ISO/TC97/SC16/N227, June 1979.~~

~~AT&T, "OPERATIONS SYSTEMS NETWORK COMMUNICATIONS PROTOCOL SPECIFICATION BX.25," Issue 2, Publication 54001, 1979.~~

~~CCIR Recommendation 476-2 (1978) and 476-3 (1982) Direct-Printing Telegraph Equipment in the Maritime Mobile Service~~

CCITT YELLOW BOOK, SEVENTH PLENARY ASSEMBLY, Geneva 1980 Volume VII, Fascicle VII. 1: Telegraph Transmission and Switching Series R and U Recommendations (Study Group IX) Volume VII, Fascicle VII.2: Telegraph and "Telematic" Services Terminal Equipment Series S and T Recommendations (Study Group VII) Volume VIII, Fascicle VIII. 1: Data Communication over the Telephone Network, Series V Recommendations (Study Group XVH) Volume VIII, Fascicle VIII.2: Data Communication Networks; Services and Facilities, Terminal Equipment and Interfaces, Recommendations X.1 - X.29 (Study Group VII) Volume VIII, Fascicle VIII.3: Data Communication Networks; Transmission, Signalling and Switching, Network Aspects, Maintenance, Administrative Arrangements; Recommendations X.40 - X.180 (Study Group VI)

~~Andrew S. Tannenbaum, COMPUTER NETWORKS, Prentice-Hall, 1981.~~

~~D.W. Davies, D.L.A. Barber, W.L. Price and C.M. Solomonides, COMPUTER NETWORKS AND THEM, PROTOCOLS, John Wiley and Sons, 1979.~~

~~D.W. Davies and D.L.A. Barber, COMMUNICATIONS NETWORKS FOR COMPUTERS, John Wiley and Sons, 1979.~~

~~Kenneth Sherman, DATA COMMUNICATIONS, A USER'S GUIDE, Reston Publishing Company, 1981~~

~~Philip F. Panter, MODULATION NOISE AND SPECTRAL ANALYSIS, McGraw-Hill, 1965~~

~~287 James Martin, TELECOMMUNICATIONS AND THE COMPUTER, Prentice-Hall, Inc., 1976~~

~~Roger L. Freeman, TELECOMMUNICATION SYSTEM ENGINEERING, John Wiley and Sons, 1980~~

~~Mischa Schwartz, INFORMATION TRANSMISSION, MODULATION AND NOISE, McGraw-Hill Company, Inc., 1959~~

~~Friend, Fike, Baker and Bellamy, UNDERSTANDING DATA COMMUNICATIONS, Texas Instruments Incorporated, 1984~~

~~William Stallings, HANDBOOK OF COMPUTER COMMUNICATIONS STANDARDS, Volume 3, Department of Defense (DOD) Protocol Standards, Macmillan Publishing Company, 1988~~

~~J. P. Martinez, G3PLX; AMTOR, AN IMPROVED RTTY SYSTEM USING A MICROPROCESSOR, Radio Communication, Radio Society of Great Britain, 1980~~

~~J. P. Martinez, G3PLX; AMTOR, AN IMPROVED ERROR-FREE RTTY SYSTEM, OZ Magazine, Copenhagen, 1980~~

~~Paul Newland, AD7I; ZAMTOR, AN AMTOR CODE CONVERTER, QST Magazine, February 1984~~

~~M. C. Lamb, N7ML; AMTOR, A HANDS-ON PRIMER, CQ Magazine, November 1983~~

~~American Radio Relay League, THE 1986 ARRL HANDBOOK FOR THE RADIO AMATEUR~~

~~ARRL, PROCEEDINGS OF THE FIRST ARRL AMATEUR RADIO COMPUTER NETWORKING CONFERENCE, October 16-17,1981~~

~~ARRL, PROCEEDINGS OF THE SECOND ARRL AMATEUR RADIO COMPUTER NETWORKING CONFERENCE, March 19, 1983~~

~~ARRL, PROCEEDINGS OF THE THIRD ARRL AMATEUR RADIO COMPUTER NETWORKING CONFERENCE, April 15, 1984~~

~~ARRL, PROCEEDINGS OF THE FOURTH ARRL AMATEUR RADIO COMPUTER NETWORKING CONFERENCE, March 30, 1985~~

~~Terry Fox WB4JFI, AX.25 AMATEUR PACKET-RADIO LINK:LAYER PROTOCOL, VERSION 2.0, ARRL, October 1984~~

~~288 David W. Borden K8MMO, and Paul L. Rinaldo W4RI; THE MAKING OF AN AMATEUR PACKET RADIO NETWORK, QST, October 1981,~~

~~Margaret Morrison KV7D, and Dan Morrison KV7B, AMATEUR PACKET RADIO: PART 1, Ham Radio Magazine; July 1983~~

Margaret Morrison KV7D, and Dan Morrison KV7B, and Lyle Johnson WA7GXD, AMATEUR PACKET RADIO: PART 2, Ham Radio Magazine, August 1983 Lyle Johnson WA7GXD, JOIN THE PACKET RADIO REVOLUTION, 73 Magazine, September 1983

~~Lyle Johnson WA7GXD, JOIN THE PACKET RADIO REVOLUTION-PART II, 73 Magazine, October 1983~~

~~Lyle Johnson WA7GXD, JOIN THE PACKET RADIO REVOLUTION-PART III, 73 Magazine, January, 1984~~

~~Margaret Morrison KV7D, and Dan Morrison KV7B, DESIGNING THE TAPR TNC AUDIO INPUT FILTER, Proceedings of the Second ARRL AMATEUR RADIO COMPUTER NETWORKING CONFERENCE, March 19, 1983~~

~~Harold E. Price NK6K, WHAT'S ALL THIS RACKET ABOUT PACKET'?, QST, July, 1985~~

~~Harold E. Price NK6K, A CLOSER LOOK AT PACKET RADIO, QST, August, 1985~~

~~289 PACTOR-Radioteletype with Memory ARQ and Data Compression~~

~~By Hans-Peter Helfert, DL6MAA and Ulrich Strate, KF4KV Gustav-Muller-Strasse 8 Lommerwiese 18 D-8948 Mlndelheim D-5330 Kobigswlnter 1 Germarry Germany~~

~~Translated by Don Moe, KE6MN/DJOHC, from the November 1990 Issue of cq-DL, published by the German Amateui Radio Club.~~

Introduction Even at poor signs{-to-noise levels usable connections car n the past ten years, Amateur Radio teleprinting has still be maintained. The rather high error rate under these increasingly evolved from an elite operating mode for conditions is tolerable in normal amateur conversations due specialists into a regular means of daily communica- to the high amount of redundancy in text. tions. Under closer scrutiny It is apparent however, that the This is a different story in the case of technical mes largest extent of progress has mainly benefited the sages, such as modification instructions, programs or the VHF/UHF operating mode packet radio. In the short wave like. In addition to the fact that every error can have disas segment, amateurs still have to make do with relatively trous effects, such texts are mostly in ASCII format. Wher modest technical standards and ease of operation. transmitted in 5-bit Baudot code, ambiguity results whict The transition from "Steam RTTY" of the T37 era to must either be tediously clarified or left to the intuition o AMTOR (SITOR) and packet radio (PR) undoubtedly the reader. represents a large qualitative improvement. However, the Recoding techniques, such as suggested in Ref 2 question may be asked whether mere adaptation of com- merely shift this problem to a different level. In addition tc mercially used transmission protocols provides a favorable significant speed reductions, transmission errors can causf solution for Amateur Radio. In the case of PR above even more disagreeable side effects. 30 MHz, this can certainly be answered with "yes." since A further disadvantage of AMTOR. which is particularly here the transmission conditions of the original system (data significant for mailbox operation, is the low effective maxi lines via telephone) are virtually identical with those found mum speed of less than 35 bauds, resulting in insufficien on VHF/UHF. In contrast, a PR connection on the 80-meter usage of the available channel capacity during phases will band in the evening often results in a serious test of good signal-to-noise ratios. patience. SITOR was especially developed for operation on short 2. Basis for the Development of PACTOR wave; but the system standard is still limited to the techni- The authors began experimenting with derivatives o cal possibilities of the 'TTL days" in the 1960s, even if the the AMTOR technology, such as longer blocks, doubling amateur designation "Microcomputer Teleprinter Over the speed, etc. Although these achieved significant im Radio" suggests something more. provements in performance, the principal weaknesses c All important AMTOR system parameters such as AMTOR, inadequate error correction and ASCII incompat character set and maximum transmission rate are geared bility, could not be overcome. Thus the design of a corn to the typical data terminal at that time, the mechanical pletely new technique was begun at the end of 1987. teleprinter. Electronic intermediate storage devices, taken Since the new system combined important characteri; for granted today, were not feasible economically. Since tics of packet radio and AMTOR, the name PACTOR wa new developments can scarcely be expected from the corn- chosen (Latin: mediator). The synchronous, half-duple mercial side, such as for nonmilitary short wave radio at basic structure of AMTOR was retained: Information block sea, radio amateurs are left to their own devices. sent at fixed time intervals are acknowledged by th The authors of this article advocate the position that receiver with brief control signals (CS). even in the era of satellite communications. efficient trans- The length of the information blocks and thus the duri mission protocols for short wave cannot be ignored. In tion of the transmission cycle is an important factor i accordance with aspects of information theory, this must determining the flexibility of a system. On a noisy channi also include the operating mode CW. long packets scarcely have a chance of survival, resultin in the PR blocking effect, which is largely minimized b 1. AMTOR-Strengths and Weaknesses Memory ARO in PT. On the other hand, during periods c Before designing a new system from scratch, it is ap- strong interference, the probability of reception increase propriate to look for possibilities of extending or improving when shorter packet lengths are used. The price in this cas existing technology. is a lower maximum data rate on a noise-free transmissic The popularity of AMTOR (see Ref 1 for functional channel. description) is well founded: the system is relatively uncom- For example: AMTOR requires a cycle length of 45 bt plicated and can be combined with available teleprinters. in order to transmit the message content of 15 bits. T% thirds of the capacity are thus swallowed up by overheat To arrive at the most favorable PACTOR block length: 'References appear on page 6. long-tern tests were performed with short wave FSK trap October 1991 290 missions and the results were evaluated by computer. The result was an optimal cycle time of nearly two seconds, which was then reduced to 1.44 seconds in the final version of the system in order to achieve short break-in times. The overhead portion now takes up only one third of the transmission capacity. In a synchronous transmission system, the stability of the clocks at both stations is very important. During the development of PT this problem was circumvented in a very simple way: the system clock at both ends was derived from the 50-Hz power grid, which provides a phase stable and Fig t-PACTOR packet format reliable synchronization framework for nearly all of western Europe. Since relying on the power grid would have naturally restricted the system's usefulness, crystal control with 3.1 Transmission Formats phase correction is now the standard mode. Due to the narrow bandwidth limitations imposed by a) Information Blocks regulations and the receiver filters, only FSK came into con- All packets have the basic structure shown in Fig 1: sideration as the modulation type. Under normal conditions, Header contains a fixed bit pattern for simplifying syn- the typical short wave channel width using a 600-Hz filter chronization and reading along, also important for permits 200 bauds at 200-Hz shift, which is achievable with Memory ARO customary filter converters without significant modifications. Data: any binary information. 192 bits at 200 bauds, 80 During periods of phase distortion, such as multipath bits at 100 bauds. propagation, which can occur on winter evenings on Status: control byte containing 2-bit packet number, break 80 meters, the speed must be reduced to 100 bauds. In or ORT request, transmission mode, etc. PT this is accomplished by optionally filling the blocks with 100- or 200-baud information. Since the main timing re- CRC: 16-bit block check code, effective on data, status mains constant, the switch over can happen automatically and CRC. without loss of synchronization as soon as requested by b) Acknowledgment Signals the receiving station. PT uses four acknowledgment signals (CS1 to CS4), As described in the previous section, the relatively high which correspond in their function to the AMTOR control probability error with AMTOR is its central weakness. There. signals except for CS4: fore PT uses a longer acknowledgement signal (12 bits) as CSI1CS2: acknowledgment function well as the cyclical error recognition code used in PR (16 normal bits), which practically eliminates the problem of unrecog- CS3: change of transmission direction (break-in), forms nized transmission errors. Only under these prerequisites packet header can this new method for amateur RTTY be fully effective CS4: requests a speed change at the sending station with the following features: The acknowledgment signals have a length of 12 bits. The characters differ in pairs in 8 bits (Hamming offset) so that Memory ARO, summing method for reconstructing the the chance of confusion is minimized. One of the most original block common causes of errors in AMTOR is the small CS On-line data compression (Huffman encoding). Hamming offset of 4 bits. Additional system characteristics: If the CS is not correctly received, the transmitting side by repeating the last packet. The request status faster and more reliable change of transmission direction reacts (break-in) can be uniquely recognized by the 2-bit packet number so that wasteful transmissions of pure RO blocks as in 100% compatible to ASCII and transmission of binary AMTOR are unnecessary. . data ORT confirmed at both stations C) Tkning independent of shift direction, no mark and space The receive pause between two blocks is 0.32 seconds. conventions After deducting the CS lengths, 0.2 seconds remain (0.17 for AMTOR) for switching and propagation optional attachment to frequency sec delays so standards such as is DCF-77 that there adequate reserve for DX operation. unique call address using complete can signs 3.2 Establishing Contact comprehensive capability for other stations to read akxq The calling station (master) sends a special synchroni- simple hardware requirements zation packet which only contains the call sign (address) comfortable operation of the called station (slave): /Header/Address (8 bytes. 100 bauds)/Address (8 bytes. 3. Important System Details 200 bauds)I Since a complete specification of the protocol would Following synchronization, the slave responds with exceed the limits of this article. the authors will provide a CS1. or CS4 if the 200-baud bit pattern was also recog- complete description in the PR mailbox network so that in- nized. Depending on channel quality the connection can terested parties can obtain further information easily. Here be started at the optimal speed without delay. The number only a few of the special aspects will be described as they of significant synchronization bytes can be chosen by the differ from the corresponding AMTORIPR procedures. slave. In practice, six characters shduld suffice. ' The OEX

291 unpleasant problem of ambiguous selcals in AMTOR is thus acknowledgment transmission in the established time eliminated. frame. This process is repeated until the sending station During the synchronization phase the relative shift has received the acknowledgment. direction is also determined. The converter or FSK setting of the two stations is irrelevant. Mark/space conventions 3.6 Data Compression can thus be dropped. In amateur RTTY only constant length characters have After receiving the first CS from the slave, the master been used to date. Frequently occurring letters such as E begins sending normal information blocks. It has proven or N require the same transmission time as X or O. useful to send system specific data automatically at the be- A frequency analysis of normal written texts shows that ginning such as master call sign, software version number the average information content per character ("Entropy" and other configuration parameter. only amounts to 4 bits, so that a normal ASCII transmission therefore contains nearly 50% ballast. Through clever 3.3 Changing Transmission Direction encoding using variable character lengths, common in Following each correctly received packet, the receiv- Morse telegraphy for the past 150 years, it should be pos- ing station can transmit a CS3 (break-in). In contrast to sible to put this idle reserve to use. AMTOR, an intermediate cycle containing no information The Huffman code (see Ref 3) used by PT approaches is not required. The CS3 forms the header of the,first infor- the optimal limit within a few percent so that nearly 100% mation packet. In the ideal situation, the direction could be increase compared to ASCII can be achieved in practice. changed again in the next cycle, which would be advan- This code can be automatically calculated as a tree struc- tageous to.- mailbox commands. Depending on whether ture based on the character frequency. The compression only the CS3 header or the entire packet was received, the occurs by packet; the character lengths vary between 2 and former master could reply with CS2 (repeat) or CS11CS3. 15 bits. Analogous to "+?" in AMTOR, the transmitting sta- A prerequisite for using data compression methods is tion can also request a break-in, which occurs by setting high data integrity. Techniques without block error recog- the BK status bit. nition such as AMTOR are therefore not suitable. 3.4 Changing Speed 3.7 Memory ARG Switching between the two speeds of 100 and 200 Customary FSK RTTY converters route the demodu- bauds is normally provided. Since an increase in speed is lated received signal via a low-pass filter to a trigger stage sensible only during good conditions and a reduction dur- where the binary data for the computer are extracted. Here ing bad conditions or a slow information flow such as manu- a weakness becomes evident: The decision whether a sig- al text entry, each direction is handled differently in the nal is a "0" or a "1" is made outside the computer and protocol. Is thus no longer accessible for "intelligent" analysis. The information whether a signal was only 1 millivolt or 1 volt a) 200 -> 100 Bauds above the threshold is lost forever. Additional distortions Following receipt of a bad 200-baud packet, the receiv- are caused by inaccurate calibration or drift of the trigger ing station can request reduction in speed with CS4. While a threshold. maintaining the time frame, the transmitting station will In recognizing for RTTY in put the packets together with 100-baud information. The the past, error techniques Amateur Radio merely the information from com- unacknowledged 200-baud information of the previous evaluated plete blocks received error-free. It is obvious that this results packet will then be repeated. In a significant waste of information particularly at poor b) 100 -> 200 Bauds signal-to-noise levels. In an ARO system such as PACTOR, A correctly received 100-baud packet could be ac- multiple repeats of an incorrectly received packet can be knowledged with CS4 which causes the transmitting sta- overlapped to recover the original information. tion to double the speed. If the following 200-baud packet The ideal solution in the form of a direct computer is not acknowledged after a predetermined number of evaluation of the audio receive signal is still constrained attempts, the speed will automatically be set back to by the high price of the necessary signal processors. Prac- 100 bauds. The decision regarding a speed change is made tice with PACTOR has shown however, that important ad- at the receiving end. Normally this occurs by automatically vances can be achieved using a low-cost 8-bd AID converter evaluating the packet statistics such as error rate. number instead of a trigger stage. of retries, number of filler characters for manual text entry, The incoming analog values are first converted to 0/1 etc. decisions as before, but additionally are stored as a sequence of 8 bit values for later use. If the 0/1 evaluation 3.S Ending a Contact does not yield packet (CRC error), the analog In the normal ARO protocol. the principle of mutual ac- values from subsequent repeats of the packet are combined knowledgment is violated at the end of a connection: One into a "sum" packet which is then subjected to a 0/1 evalu- station sends the appropriate ORT signal and then switches ation and CRC check. the transmitter off. If the other station does not receive the There are two potential problems to be considered: in- ORT signal, it sends acknowledgment signals until the a) If the acknowledgement signal sent by the receiving sta- ternal timer expires, which leads to undesirable interfer- tion is not immediately heard, old RO packets could be ence, particularly in AMTOR mailbox operation. added to the sum for the new block. PT solves this problem through a special mode: At the end of a connection special ORT synchronization packets b) Packets which were nearly completely destroyed, such are transmitted which contain the receiver address in the as by an interfering carrier, should also be ignored. reversed order. If the slave station recognizes such a packet To handle these cases, the information contained in during the normal search phase. it responds with a single the packet header is evaluated. This is inverted in phase October 1991

292 system since a home computer solution (C64 or similar) require- 100 nearly always has disadvantages such as space ments, HF interference, etc. . ... with Memory- ARO The "PACTOR Controller" (PTC) was implemented as so without Memory- ARQ a single board computer based on the Z80 processor. Ini- tially DL6MAA built a prototype with the new CMOS SMD P 60 PACTOR, 200 Boud. i; chip TMPZ84COl5 which contains all of the necessary Z80 z 5/H standardized to bandwidth of 600 Hz. peripheral modules. At the same time a compatible version h- (Gaussian white noise) was built by DF4KV using conventional Z80 components. A battery-powered real-time clock is integrated along with two EIA-232 interfaces as terminal connections (XONIXOFF handshaking). The system status is displayed 0 L., J ... I--T.*1*' *; I I 1 i I in a block of 12 LEDs, and an additional row of LEDs acts -20 -15 -10 -s 0 s 10 1s as a tuning indicator based on the incoming analog values. S/H (d8) Important events such as connect or ORT are also signaled acoustically. The associated converter operates with filters. The output signal from the low-pass filter is fed via a level Fig 2-Increase in data rate when using Memory ARQ. converter stage to the ADC input. The software was developed primarily by DL6MAA and in addition to the pure PT system also contains a command The free RAM with the packet counter in order to guarantee a simple re- interpreter with TNC compatible syntax. quest recognition. Obviously unusable packets, such as memory is automatically configured as a private mailbox deposited or read. those with a constant bit pattern or a destroyed header are in which messages from callers can be not added to the summation. The available program memory is large enough to accom- Similar methods are used in professional transmission modate additional software such as AMTOR or a CW keyer. PTC on 80 technology.4 The term "Memory ARO" comes from there. In the fall of 1989, DL2FAK installed a Sample calculations confirmed in practice have shown meters with access to the message base of a 70-cm PR that the data rate can be significantly improved at poor sig- system. Using this PT link technique, ASCII compatibility nal levels by using Memory ARQ. It is particularly appar- and speed of the PT system were demonstrated very well. sharp PR limit at which the communication In the meantime, further amateurs have become in- ent that the who collapses has been eliminated. (See Fig 2.) volved with PT tests: DK5FH, DF4WC, and DL3FCJ, Memory ARO can be limited to being used only if the has developed a new PTC design with an integrated error recognition technique has adequate reserve. In converter. dramatic increase in the rate of incorrect charac- AMTOR a 4.2 Operating Techniques ters would result. For users already familiar with packet radio, there is 3.8 Monitor Mode scarcely any adjustment needed. A series of commands The requirement that other stations be able to read the allows checking and setting all important system communications is specified in the Amateur parameters. Most commands can also be accessed by the content of mes- Radio regulations. This is necessary when there are more remote station for reading directory contents, storing than two active stations. PACTOR is similar here to the sages, etc. monitor mode in PR. The received bit stream Is continu- checked for valid packets according .5. Conclusion ously evaluated and project to CRC. Since several samples occur per bit and a separate PACTOR was developed as a purely amateur Error CRC must be calculated for the two possible speeds. the in the context of the experimental radio service. of the Z80 CPU is nearly at its limit in this mode. robustness and short wave suitability were primary goals. capacity be- The result is an uncomplicated read-along operation, which PTC kits based on the design of DL3FCJ should intervention, in contrast to the L- come available by the end of 1990. Interested parties can works without operator an mode in AMTOR. request circuit diagrams from the authors by sending Additionally an automatic CW identification occurs in SASE. Programmed EPROMs and GALS are likewise avail- 5-minute intervals. able in limited quantity. 3.9 Supervisor Mode 6. References The definition of special control characters makes It 1L. Monz. DFSPD: Funkfemschreibverfahren. cq-DL 1186. possible to pass information from the receive buffer direct- (Radio teleprinter techniques) ly to the system level. Through this supervisor mode. vari- =A. Clemmetsen, G3VZJ: 10 Jahre AMTOR-Rkkblick and ous protocol versions could be automatically harmonized Ausbtick. cq-DL 2189 (10 years AMTOR-review and and thus future developments of the system could be taken prognosis) into consideration. 3F. Topsoe: Informationstheorie. Stuttgart 1974. (information 4Jtheory) Practical Operation 4. Me) r, D. Chang: Efficient Selective Repeat ARO 4.1 Hardware and Software Strategies for Very Noisy and Fluctuating Channels. IEEE PT was designed from the start to be a self-contained Transactions on Communications, vol. COM-33, May 1985.

~~OEX~~

~~293 PTC-The PACTOR Controller~~

~~By Martin Clas, DLIZAM and Peter Mack, DL3FCJ Krouzbergstrasse 24 Im Bangert 21 D-6457 Maintal 4 D-6450 Hanau 1 Germany Germany~~

~~Translated by Don Moe, KE6MNIWOHC, from the July 1991 Issue of cq-DL, published by the German Amateur Radio Club.~~

In the November 1990 Issue of cq-DL (see English transle- don beginning on page 3 of this issue of OEX) DF4KV and DL6MAA described the new radio teleprinter technique, that they had developed, called PACTOR and they announced that suitable hardware would soon become available. In the following article this hardware, the PACTOR Controller (PTC) TYNE OLIZAM PACTOR Is extensively described. The currently available software ® fa version, 1.0, also supports complete AMTOR and RTTY HM09" Tratfic *ASCII operation. aoM NN HYftnwa he PTC is implemented T on two circuit boards. The CHO Raowet AYTOR main board is Euro-Card size (160 x 100 mm) and oarrer ORTTY contains the Z80 processor and its support compo- A r.A nents as well as all interfaces. The second circuit board contains the display field, consisting of 20 variously colored, low power, light-emitting diodes. This display board is sol- Sample unit of the PACTOR controller dered at right angles to the front of the main board. Double- sided circuit boards are used that have plated-through holes. a silk-screened component layout and solder resist, The processor clock is provided by inverter U11A thus fulfilling requirements for high quality. The hardware together with crystal G1 and Ct, TC1 and R5. The exact is kept as simple as possible for equipment of this com- frequency can be set with TC1. The oscillator signal (C7) plexity. The entire assembly is mounted in a box of alumi- is fed via buffer U11B to the processor. num profile similar to that used by the TNC2c or the AMC. A MAX691 (1.110) was chosen to provide a reliable On the back panel are the connectors: a 25on sub-D power-on reset and backup power from a lithium battery. for the EIA-232 connection to the terminal, a 5-pin DIN for In case of loss of operating voltage, it maintains the RAM the connection to the transceiver, and a third jack for power. (1.13) and the real-time clock (1,18). The time and date will The PTC is thus fully compatible with the connections on therefore not be lost whenever power is off or a reset is the TNC2c or the AMC. The holes and cutouts on the front performed. and back panels are stamped out. The front panel has silk- U10 also performs the "watchdog" function. The screened labels. proper operation of the software is watched over bythe WDI The PTC can be powered from a do source between input (pin 11). If the pulses on WDI cease, U10 generates 9 and 14 V and requires approximately 200 mA. a reset and the program is restarted. Thus a program crash will not lead to a long term failure of the system, but only The Circuit Processor Section to a brief interruption. This is a major criterion for an The central unit of the processor section consists of independently operating system. the 8-MHz CMOS version of the well-known ZOO proces- Components U6, U7 and U13 drive the light-emitting sor (1.11) and a multipurpose component, STI (114), which diodes. U6 and U7 are merely latches. and U13 decodes contains serial and parallel ports as well as four indepen- three data lines to drive the tuning indicator, which is con- dently programmable timers. These are used to generate trolled directly by the program. It displays what the proces- the AFSK and FSK signals as well as the clock frequency sor has processed from the input signal. LEDs not needed for the programmable filter (U15). The STI also contains the for tuning are switched off by the software so that the user EIA-232 interface and a MAX233 (US) converts the signals sees a better tuning indicator. to/f rom the standard voltage levels. The port can be oper- The PTC contains an analog-digital converter (U12) for ated at 300, 1200, 4800 or 9600 baud, 8 bits, one stop bit transforming the preprocessed audio input signal from the and no parity. radio receiver into the digital signals required by the proces- The program is stored in a 27C256 EPROM or 27P256 sor. This function is necessary for the Memory ARQ, as PROM. A battery-powered CMOS RAM (U3) maintains the described in Ref 1. Transistor T3 assures proper initializa- data and settings should the operating voltage be removed. tion of the AID converter following a reset. The conversion October 1991

~~294 Schematic of analog section~~

~~'Y VV! I 2d~~

~~H -d U f__Hi 0~~

~~N d~~

~~P T~~

~~w ,-w ..- i-~~

~~M. .1 II; r~~

~~a is~~

~~jf i i is~~

~~Circuit board of the PACTOR controller~~

messages with 'Y/W", to read messages with "UR". to display the contents with "11DIR", etc. Commands with two slashes are remote control commands. and without a double-slash. these commands control the mailbox in the local PTC. Naturally the contents of the mailbox are maintained by a battery in case of power loss. The size of the mailbox memory is 21,518 bytes. In version 1.0 of the PACTOR software, the mailbox is only accessible in the PACTOR mode. Remote control of the local station can also be disabled. The PTC keeps its own logbook which can be displayed with the command "LOG". This command is also available to remote stations so that it's possible to view the log of a remote station. Only the most recent 16 -contacts are stored in the logbook, which is likewise maintained by the battery. As already mentioned, the software also supports alignment of the PTC. Only four alignment points are provided: two in the demodulator, the audio level to the microphone input of the transceiver. and the crystal frequency. If the command "ME" is entered, a message appears requesting

295 speed is determined by C21 and R18. Since the PTC is command operated, it displays th Address decoding for the various components of the prompt "cmd:" and waits for input. All commands are tee PTC are performed by a programmed logic component, minated with a CR. Corrections can be made using th GAL 16V8 (U9). The GAL provides the chip select signals backspace key. In the stand-by mode, the PTC is immed on its output. It is especially programmed for the PTC and ately ready for another command. thus cannot be substituted. When establishing a connection and in the connects, state, characters arriving over the EIA-232 interface Audio Processing of the Received Signal are ser directly to the transmit buffer to be transmitted at the nex The demodulator operates as an envelope detector opportunity. unless preceded by Escape. The size of th (filter converter), but differs from customary circuits of this text buffer is set to 5000 characters. During a connectior type due to the sharp cutoff characteristics of the input all PTC commands must be prefixed with Escape, and onl bandpass and the optimization of the low-pass fitter at base- one command can be entered at a time. In case of an err: for band 200 bauds. in the command, PTC accepts an immediate reentry of th, U1 8B and U1 SC form an active fourth-order high-pass- command. U15 functions as an elliptical seventh-order low-pass with switched capacitances (SC filter) and raises the filter func- The Command Structure tion to a sharp cutoff band-pass. The SC technique permits Some commands expect arguments, others not. At the processor to adjust the limit frequencies via FCIk. per- argument must be separated from the command by at lea: mitting quick adaptation to the baud rate. U15's other one space. If a command expecting an argument is entere advantages are its very small size and excellent long-term without any argument, the current setting is displayed. characteristics. Nearly all commands can be drastically abbreviates U1 8A amplifies the preselected signal into M*ng. The for example. the command "Call" can be shortened to "C" actual discriminator filters are U1 7A and U178. Their out- Characters following the minimal length are ignored up t( put signals are rectified by V5 and V6 and are combined the first space. Since describing the complete PTC corn in a summation low-pass (U17C). Before arriving at the AID mand set would exceed the limits of this article, anyon( input, the freshly demodulated signal passes through Interested in building or acquiring a PTC should contact tht another fourth-order low-pass, which is designed to achieve address provided at the end of this article, where he cai optimal immunity from noise or interference at a 200-baud obtain complete information, components and naturally. data rate. handbook. The characteristic curare of the demodulator indicates Several PTC commands are comparable to those it a sine-wave dependence between the input frequency and a TNC so that anyone with packet radio experience shout( the output voltage. have no trouble operating a PTC. For example, the com mand Is A/FSK Transmit Signal "C" "Call" or "Connect". "D" is "Disconnect" and "MY" is "MyCall" which changes the call sign store( The tones to be transmitted by the transceiver are in ROM. generated directly by component U4 (STI). Since it can only supply square-wave output signals. filtering by the active Special PTC Features low-pass U18D is required before the tone signals arrive at the microphone input of the transmitter. An Intelligent piece of equipment such as the PTC Is After removal of the harmonics, the signal passes capable of operating more than just PACTOR. It can also through P3 and C39 to the interface to the transceiver (X3). support the operating modes of AMTOR in all its associat- The AFSK tone frequencies are 1200 Hz and 1400 Hz. In ed variations. such as ARQ, FEC, list, etc, and RTTY. at the case of a transmitter with a direct FSK input. FSK out- all speeds between 30 and 300 bauds. put from the PTC is provided by transistor T2. When the command "AM" is entered, PTC switches The user sets jumper BR8 to chose whether FSK or to the AMTOR mode. The selcal stored in ROM then ap- AFSK should appear at jack X3. pears on the screen of the terminal and another selcal can be entered at any time. The AMTOR and PACTOR calls Operation of the PTC remain in memory until the restart command is executed. When power is applied, the PTC performs a test of the Restart resets all settings to those in ROM and completely LEDs and sends a sign-on message containing the call sign erases the RAM. stored in ROM to the terminal. The command "BAU < n > " switches to RTTY at the OEX 296 ffim it in In 16 the the the fre- the the that con- com- in by of control so

~~the bytes. the battery recent crystal requesting control a with section the remote by stations most~~

~~display 21,518 alignment accessible demodulator,~~

maintained and to are is the appears disabled. analog the only be remote in of displayed commands is to Only "I/R", supports maintained be likewise two slashes memory also Is message these can with are also transceiver, a two mailbox can station.

~~Schematic the the available mailbox which provided: with which of~~

mailbox entered, station also remote the software are is messages is a input the of the double-slash, of local logbook logbook, of software, read a points size log "ME" Commands to the own the PTC. The of its the in etc. command

contents microphone without PACTOR mentioned, "//W", local loss. This view alignment the command the the keeps stored and control to to the of four with the "//DIR", mode. are PTC If in power 1.0 already level "LOG". Only with of As Naturally Remote The possible 298

PACTOR battery. PTC. audio quency. messages mailbox case version mand it's tents commands, -contacts that the alignment point Pt should be set to the maximum and "Error" illuminate simultaneously, data packets are value displayed on the screen. Following "return." PTC dis- being reconstructed using Memory ARO. plays numbers on the screen and P1 should be adjusted One LED shows if the PTC is in connected mode. This to the maximum value. P2 is then set in the same manner. LED flashes it a connect has taken place when the opera- At the same time, the AFSK output produces audio for setting tor was not present, such as when a message was written the audio level to the microphone input of the transceiver. into the mailbox. In this article, the PACTOR operating The setting of the crystal's frequency is very important mode was not discussed in greater detail since this has for synchronous operating modes such as PACTOR and already -been done in Ref 1. AMTOR. It is adjusted with TCt in conjunction with a fre- Anyone interested in building or buying a PTC should quency counter. If the frequency is not set correctly, the contact (include an SASE): Dr. Thomas Rink, DL2FAK, software will automatically compensate for the offset dur- Fl6ntgenstrasse 36, D-6450 Hanau 1, Germany or the ing a PACTOR contact. This is naturally only possible within authors of this article. limits a proper adjustment should definitely certain so that References be performed. ' Using the command "Phase", the amount of compen- IDL6MAA, DF4KV: PACTOR - Funkfernschreiben mt sation can be displayed during a contact, also remotely. Memory-ARO and Datenkompression. cq-DL 11190 with Memory ARO This opens the possibility of setting the frequency by com- (Radio Teletype and Data Com parison to a correctly aligned reference station. Alignment pression) is therefore possible without any test equipment. Alternatively, an external reference clock source can also be attached. The "Show" command, followed by various arguments, displays the settings in the PTC system. The command is also usable from a remote station. so that the system parameters can be copied to the remote station. Among other things. the Reprinted with permission from QEX, command can be used toredisplay the last 1.600 characters received. American Radio Relay League Thus text can be brought back that has already scrolled off the screen. which would be very helpful on simple terminals or small displays. LED Status Display The PTC contains a very practical tuning and status display with 20 LEDs. Eight LEDs are in the tuning indicator and the remaining twelve are in the status display. It can thus be seen who has the keys and whether a changeover takes place. The changeover (CHO) can be I initiated at any time by the transmitting station. The character for it can be freely chosen, but is normally set to Ctrl-Y. The receiving station also has the chance to force a CHO. The display also indicates whether Huffman data compression is in operation or whether 8-bd ASCII is being transmMed. The latter is useful when text with IBM graphic characters or the like are to be sent. The PTC adapts the transmission speed to the channel conditions. thus one LED displays the speed, 100 or 200. The switch-over criteria can be altered with special commands. Ad- ditionally, the indicators show whether data transmission is occurring. or if there are currently idle, request or error conditions. or whether AMTOR or RTTY are active. It the t.EDs "Traffic" `299 CLOVER is a term you might have heard about in the past. If not, you certainly will be hearing about it in the future. K9GWT and W7GHM give us a rundown on this new form of HF data transmission. CLOVER Fast Data on H F Radio

~~BY BILL HENRY", K9GWT, AND RAY PETIT', W7GHM~~

If you operate RTTY, AMTOR, or packet, you've probably seen a few references on your screen to something called "CLOVER." "What is it and why do I need it?" is the usual reaction. Yes, "clover" (little letters) Is a plant, sometimes "wished over" (and sung about by Arthur Godfrey). But "CLOVER" (big letters) is a new way to send data on HF radio that Ray Petit, W7GHM, has invented. This is the story of CLOVER, a project that continues to this date.

~~What Is CLOVER?~~

CLOVER had its beginnings about 15 years ago when Ray and others were experimenting with very narrow bandwidth Morse code. It's called "Coherent CW."' When packet radio came along, Ray tried packet on VHF and then HF. As most of us have found. HF packet radio leaves a lot to be desired. The Ionosphere is just not very kind to packet data, and often many repeats are required to pass any data at all on 20 meters. Unlike the rest of us, Ray quickly realized that putting "bandaids" on HF packet or AMTOR was just not going to do the trick; what was needed was a new approach. The new approach must be bas- Fig. 1- Comparison of AMTOR, Hr= Packet, and CLOVER41 spectra. ed on a thorough analysis of the real HF signal conditions and on techniques that can compensate for these conditions. Ray started by just listening to real radio leaf." In July 1990 Ray published the first breath of fresh air to us. Finally someone signals, observing their fades and phase paper describing the mode in 00(.2 The had taken the pains to start at ground zero changes on typical HF paths and under name "Cloverleaf" came from the obser- and devise a modulation format that would varying conditions; shortwave broadcast vation of a scope pattern while watching work on HF. Very quickly Ray and HAL signals made great "test signals." Com- the data; it was a perfect four-leaf clover. teamed up to continue work on his new bining information from signal observa- As Ray's work continued, the pretty scope "CLOVER Modulation." tions, reading, and previous work on Co- pattern was lost to technology, but the Unique features of Cloverleaf include (1) herent CW, Ray devised a new way to send shortened name "CLOVER" remains. multilevel phase modulation, not FSK; (2) data on HF radio which he called "Clover- Other amateurs had also been search- use of sequential pulses whose state ing for a way to cure the problems we were changed only between pulses (not wherta having with sending data on HF. Bill Henry, carrier is on the air); (3) very low base data K9GWT, and Jim Tolar, W8KOB, of HAL rates (25 bps); and (4) very tightly con- HAL Communications, P.O. Box 365, had also been working on the problem for trolled frequency spectra with no side- Urban, IL 61801 several years. Ray's paper was like a bands (100 Hz total bandwidth to - 60 dB).

300 Cloverleaf could pass error-corrected data was also limited, since several "routine" of data link-"ARO mode" (ARQ stands over a typical HF path about two to three processes had to be moved f rom the 6809 for Automatic Repeat Request). times faster than AMTOR or HF packet ra- to the DSP56001. AMTOR characters are coded so that dio. Unfortunately, Cloverleaf also made In spite of these limitations, CLOVER-11 the receiving station can detect an error extreme demands on the HF radio equip- in.SUMMER CLOVER hardware worked in each character sent. The sending sta- ment. Radio frequency accuracy and sta- very well. Ray devised an adaptive ARO tion sends three characters, turns his bility had to be of the order of t 0.1 Hz! mode in which the modulation parameters transmitter OFF. and listens for a one- This is way beyond the capabilities of any are automatically adjusted to fit ionosphere character response from the receiving sta- currently available commercial radio conditions. We were able to put CLOVER tion. The response is either "all OK, send equipment. Ray also designed a,complete on the air and run several tests. We also next three," or "repeat last three charac- transceiver to use Cloverleaf modulation.' ran extensive laboratory tests under a ters." By this means AMTOR offers error At the time when Ray and HAL first number of different simulated conditions. correction. However, like RTTY, it is-also teamed up, Ray had already started work Our avowed goal of increasing data speed "slow." Under the best of conditions AM_ to include new DSP (Digital Signal Pro- to ten times faster than HF packet or AM- TOR can pass data at an equivalent RTTY cessing) technology in his Cloverleaf mo- TOR was easily met. As always happens rate of 50 baud (6.67 characters per se- dem. DSP offered many advantages over in an R&D project, we also found a number cond). AMTOR is also limited to the same the basically analog Cloverleaf circuitry, of new areas to consider-and some character set as Baudot-all capital let. the major ones being (1) greatly reduced features that needed improvement. ters and no ASCII control characters. radio stability and accuracy requirements In November and December 1991 we Because of the efforts of Vic Poor, (to t 10 Hz), and (2) faster data throughput went back to the drawing board. Ray and W5SMM, and his APlink network program, (to 750 bps). Ray and I soon decided to put I AMTOR has seen a resurgence of interest ..all both knew that we needed more micro. our eggs in the DSP basket." The origi- processor horsepower to do all that we over the past three years. AMTOR nets nal "Cloverleaf" modem was renamed wanted. We also realized that CLOVER have also pioneered the use of frequency "CLOVER-I" and the new DSP version would never be a success if each unit had scanning radios to allow a bulletin board dubbed "CLOVER-II." to sell for $30001 What has evolved is still station (BBS) to serve many users at vary- Ray and HAL worked on development another hardware- and software version ing distances on different bands and of CLOVER-II all through the fall of 1990 which we call "PC-CLOVER." As the name frequencies. and spring of 1991! The first working implies, PC-CLOVER is a plug4n card for H F packet radio is an out-growth of VHF CLOVER41 modems were demonstrated at IBM-compatible personal computers. packet radio, pioneered by the Tucson Am- the Dayton Hamvention in April 1991. The While we can't do much about the present ateur Packet Radio Corp. (rAPR). Like AM- new modem had a bandwidth of 500 Hz (to high cost of DSP technology, PC CLOVER TOR, packet radio (AX.25) uses an ARO. - 60 dB), passed error-corrected data at does not need expensive cabinetry, power type of format to automatically sense er- rates up to 750 bps, and would work with supply, and front-panel hardware, but does rors and request repeats. However, pac- "normal" HF transmitters and receivers. include the much needed additional pro- ket radio supports the full 7-bit ASCII char- The Dayton demonstration equipment was cessing power, primarily a 68000 IC in- acter set, including upper/lower-case let- admittedly "primitive" and there were stead of the 6809. These changes have re- ters and control codes. VHF packet radio many details yet to be worked out. How- duced the price dramatically. works very well and has become the deface ever, Ray and 1 wanted to show it and see As this article is being written (early Feb to VHF mode for data transmission. If there were any other amateurs who were ruary 1992) PC-CLOVER development is Many aspects of packet radio, however, interested. We were convinced that we well underway. A number of very thorny conspire to make its performance on HF had a marvelous machine, but the inven- problems have already been licked. The very disappointing. The major problems tor can easily fall In love with his gadget, first public exhibition of PC-CLOVER will be with HF packet radio are (1) the modula- even If there is no market. at the 1992 Dayton Hamvention. tion format (300 baud, 200 Hz shift FSK), The results of the Dayton showing were (2) the AX25 protocol (long blocks with on. beyond our wildest dreams. Everyone who y CRC error detection and large amount sawCLOVER4I work was impressed-and How CLOVER Works of overhead), and (3) the wide bandwidth wanted one or two! required in today's crowded HF bands (2 kHz). Under perfect ionospheric As a result of comments from those at To adequately explain why we feel CLOVER conditions HF packet radio could Dayton and from new ideas that developed is such a breakthrough, we must first brief- send data at as CLOVER-11 was prepared for the show, ly review the pluses and minuses of exist- up to 20 ASCII characters-per-second. However, what happens in we decided to build a new "universal" ing HF data modes-RTTY, AMTOR, and fact Is that typi- hardware base for development work, HF packet radio. cal HF packet data is passed at only a rate quickly dubbed "SUMMER CLOVER." RTTY of course led the way for "auto- of 4 to 6 characters-per-second (about the HAL built a total of eight such units. Like matic" reception of characters or data via same speed as AMTOR), and a HF packet Ray's original DSP design, SUMMER HF radio. RTTY has been around since the signal requires twice the bandwidth of an CLOVER units used Motorola DSP56001 1940s and is very reliable. The techniques AMTOR signal. and 681309 processors. We had hoped that we use today to send and receive RTTY are Since all VHF traffic networks now use SUMMER CLOVER would meet all of our much the same as those first used. We packet radio, HF packet radio networks requirements and that we could build many have better equipment, but use the same have evolved to provide long-distance sup- of these units for use in "beta4esting" of FSK modulation and Baudot or ASCI I code. port. Pioneering work has been done by HF the new mode. RTTY is slow and does not offer error cor- packet stations participating in the ARRL- Enter Murphy and his infamous law! rection. RTTY speeds of 60 WPM (45 baud) sponsored "HF Packet STA" program. First, SUMMER CLOVER hardware was to 100 WPM (75 baud) are common. In- These fellows have invested a lot of time, extremely expensive-about $3000 each. creasing the RTTY speed increases the money, and persistence in making HF pac- We could not afford to build a lot of them probability of errors; we generally use 45 ket work. to be loaned out for "beta-testing." Sec- baud. CLOVER intends to support the many ond, our "bright ideas" for features soon AMTOR evolved from an existing ship- advantages of AMTOR and HF packet ra- out-grew the capabilities of our hardware! to-shore "radio telex" mode, often called dio and "fix" the major problems of these Ray's software soon consumed the entire "TOR" or "SITOR" (CCIR 476 and CCIR modes. The most serious limitation of capacity of the 6809. DSP performance 625). AMTOR introduced us to a new type RTTY, AMTOR, and HF packet is data

301 throughput and how the data is used to CLOVER uses different modulation tech- ulation practical. Phase reference deter- modulate the radio signal. The ionosphere niques. First, CLOVER shifts the phase and mination, phase detection, and pulse am- is not a "friendly" medium for data signals. not the frequency of the carrier. Second, plitude shaping are all tasks performed HF signals often arrive at the receiving an- more than one bit of data can be sent per very rapidly by the DSP. However, CLOVER tenna by many different propagation paths; phase state. For example, BPSK (binary modulation is sensitive to phase inaccura- two or more paths are common. Each sig- phase shift keying) has two phase states cy (or "dispersion"). To sense 16PSK lev- nal path has its own time delay, amplitude, (0 or 180 degrees) which can be used to els, we must be able to detect phase and even different center frequency. The represent MARK and SPACE. OPSK changes of ±22.5 degrees and be receiving antenna does not discriminate; (Ouadrature PSK) has four phase states (0, synchronized to the transmitted signal to it adds all signals and passes the compos- 90,180, and 270 degrees). Asingle phase within t 12.25 degrees. Since the iono- ite on to the receiver. The amplitudes and change in OPSK represents the state of sphere adds phase "dispersion," a good phases of the separate AC signals com- two binarybits of data. Similarly, 8PSK can stable signal-and lots of DSP process- bine algebraically to produce a widely vary- send the state of 3 bits per phase change ing-is required to make this measure- ing receiver input. Deep selective fades and 16PSK can send 4 bit states per phase ment. As CLOVER progresses from BPSK and time-smearing of data pulse transi- change. to 16PSK to increase data throughput, in- tions are the usual result. CLOVER also allows use of Amplitude creasingly better signals are required. Once combined at the antenna, the indi- Shift Keying (ASK) in the 8PSK and 16PSK However, when signals are good, CLOVER vidual path signals are not easily separated. modes. We call these modes "8P2A" (4 takes full advantage and really "moves the It is usually impossible to compensate for data bits per phase/amplitude change) and bits." all of these "multipath" effects in the de- "16P4A" (6 bits per phase/amplitude CLOVER also takes a different approach modulator. A good example of multipath change). Since all changes in phase or to error correction. AMTOR and packet ra- ionosphere distortion is the "selective fad- amplitude occur at the fixed base rate of dio both correct errors by sensing errors ing" we hear when listening to music from 3125 bps (an equivalent pulse width of 32 at the receiver and then requesting repeat a shortwave radio station. While annoying ms), data errors due to multipath time transmissions. When there are errors to be when listening to music, this distortion can smearing of data transitions are minimized. fixed, data throughput is slowed by the time be totally destructive to data transmissions- The CLOVER modulation "strategy" is it takes to send the repeats. When condi- A major nonrecoverable parameter of to always send data at a very slow base tions are poor, packet radio often bogs HF data is the time at which the data state modulation rate and to use multi-level down, sending only repeats and no data; changes from MARK to SPACE, the data chariges in phase or amplitude to speed- AMTOR will slow-down considerably under transition time. If we lose this information, up data flow. One final twist to CLOVER-1I the same conditions. the modem can no longer tell when one is that there are fourseparate transmitted CLOVER uses a Reed-Solomon error- data pulse ends and the next begins or if the pulses, each separated by 125 Hz. Each correction codes which allows the receiv- logic state should be a "1 " or a "0." When of the four pulses may be modulated by er to actually fix a limited number of errors two signals arrive with different propagation BPSK through 16PSK plus 8P2A or 16P4A without requiring repeat transmissions. time delays, the composite antenna output modulation. This further multiplies the ef- For a moderate number of errors, CLOVER signal is "smeared" and the transition fective data throughput by a factor of four. does not require repeats and data contin- times overlap. Measurements by Ray and Putting it all together, CLOVER can send ues flowing at the no-error rate. To dis- many others show that we can expect this data at rates from its base data rate (31.25 tinguish between the two schemes, we time overlap from different paths to be as bps) to 24 times its base data rate (750 classify AMTOR and packet radio as "er- much as 3 to 5 milliseconds (ms). Typical bps). Wow! It's almost like something for ror-detection" protocols and CLOVER as demodulators (and UARTs) must receive nothing! Not so by a long shot. There are an "error-correction" protocol. In addition, at least one half of each data pulse without still problems to be solved! like packet radio, CLOVER includes a CRC distortion to determine the MARK or PSK modulation Itself poses some pretty (Cyclic Redundancy Check sum) which is SPACE data state. Therefore, the narrow- serious problems. If we modulate a contin- used when conditions are very bad and the est data pulse which can be reliably de- uous carrier using PSK, the frequency number of errors exceeds the capacity of modulated is on the order of 6 to 10 ms, spectrum we get is very bad for HF use, the Reed-Solomon error corrector. corresponding to maximum data rates in as sidebands are strong and extend over CLOVER ARO mode is also adaptive. As the range of 100 to 167 baud. Observation a wide spectra. CLOVER avoids this prob- a result of the DSP calculations necessary shows that the 100 baud limit is more real- lem by twotechniques: (1) each of the four to detect multi-level PSK and ASK, the istic and even it can be too high for satis- tones is an ON/OFF amplitude pulse and CLOVER receiver already has information factory data transmission at times. the phase is changed only when the pulse which can be used to determine the signal- HF packet radio uses a 300 baud data is OFF; (2) the amplitude waveform of each to-noise ratio (S/N), phase dispersion, and rate, a pulse width of 3.3 ms. Successful ON/OFF pulse is carefully shaped to mini- time dispersion of the received signal. HF packet transmissions are therefore mize the resulting'frequency spectra. CLOVER has 8 different modulation very unlikely if the signal is propagated by Combined, these techniques produce a modes, 4 different error correction set- multiple paths. H F packet works well only composite CLOVER spectra that is only tings, and 4 different data block lengths when the operating frequency is close to 500 Hz wide down to - 60 dB. This is one which can be used-a total of 128 different the Maximum Usable Frequency(MUF)- half the radio bandwidth required for AM- modulation/code/block combinations. when there is only one propagation path. TOR and one quarter that for HF packet Using real-time signal analysis, the Since this is the exception and not the rule, radio. A comparison of AMTOR, HF pac- CLOVER receiver will automatically signal long-term packet performance on a single ket, and CLOVER spectra is shown in fig. 1. the transmitting station to change modes fixed frequency is pretty poor, and many Detecting PSK is a lot more difficult than to match existing ionosphere conditions. repeats may be required to pass any data detecting FSK. We need a very accurate When propagation is very good, CLOVER at all. phase reference to determine which phase can set itself to the highest speed and data HF packet radio, AMTOR, and RTTY all state is being received. Analog phase de- literally "screams" down the path. When use FSK modulation. One radio frequen- tection and PSK recovery circuits can be conditions are not so great, the data speed cy is sent for the "1 " or MARK pulse state very complicated and expensive. Fortu- is slowed. As noted earlier, the CLOVER and another for the "0" or SPACE state. nately, the microprocessor and now DSP character throughput rate under typical The transmitter carrier frequency is shifted have greatly simplified the task. HF conditions is about ten times faster back and forth at the same rate as the data. DSP is the keyto making CLOVER mod- than AMTOR or HF packet. However,

302 CO, June 1977, pp. when we get one of those "perfect iono- mode has evolved from the need to pass Weiss, K8EEG/0, 24- 48-54. "Coherent CW,'- sphere" conditions, CLOVER will "shift data via HF radio at a faster rate and from 30; July 1977, pp. W6NEY, OST, May gears" and pass data at 50 to 100 times an observation of the real-world propaga- Charles Woodson, June 1981, pp. 18-23. the speed of AMTOR or HF packet radio. tion conditions. It answers a pressing need 1981, pp. 11-14; 2. 'The CLOVERLEAF Performance-Or- In all cases, CLOVER automatically chang- to send data more reliably and faster than iented HF Data Communication System,', es speeds to give the maximum speed that can be done using AMTOR or HF packet Ray Petit, OEK, July 1990, pp. the ionosphere will allow. radio. CLOVER is admittedly a very compli- W7GHM, cated mode that has only recently become 9-12; reprinted in ARRL/CRRL 9th Com- practical due to the advent of relatively puter Networking Conference Proceed- Is CLOVER Legal low-cost DSP devices. CLOVER is also ings (1990), pp. 191-194. For Amateur Use? very bandwidth efficient, requiring a small 3. "Frequency-Stable Narrowband fraction of the spectra of AMTOR (one half) Transceiver for 10100.5 kHz," Ray Petit. We hear this question often. The short an- or HF packet radio (one quarter). W7GHM, ARRUCRRL 9th Computer Net- swer is yes. The reason lies in the defini- Although bandwidth efficiency may not at working Conference Proceedings (1990), tion of the CCIR Emission Designator' and present be high on the amateur's list of pp. 191-194. how that matches our FCC Part 97 Rules "must haves," we must realize that while 4. "CLOVER-11: ATechnical Overview," and Regulations. As can be seen in fig. 1. amateur radio itself is growing, our HF Ray Petit,W7GHM,ARRLArriateurRadio CLOVER bandwidth is 500 Hz-no doubt f requency allocations are likely to remain 10th Computer Networking Conference about itI Sin,:e the CLOVER modulator fixed. In the future, we must find ways to Proceedings (1991), pp. 125-129. generates tones which drive an LSB trans- cram more signals into our available HF 5. Reed-Solomon encoding modifies mitter, the modulation mode is "J2." One spectrum. Like SSB versus AM, CLOVER's transmitted data in a pattern that the re- possible point of oonfusion: While CLOVER bandwidth reduction allows us to make ceive modem uses to detect and correct does use multiple tones and multiple mod- more efficient use of the limited H F bands errors caused by ionospheric distortion. ulation levels, is not a multiplex we have. CLOVER is still evolving as this Transmitting and receiving CLOVER mo- emission; we are sending only one data article is being written. dems are synchronized so that original bit stream over the air. The full CCIR emission patterns are restored when receive data designator for CLOVER is "500HJ2DEN." is processed and passed to the data ter- Thisall agrees with FCC Part 97 Rules and Footnotes minal. This type of "Forward Error Correc- Regulations. tion" (FEC) allows correction of errors 1. For more information about Coherent without requiring repeat transmissions. CW, see "Coherent CW-Arnateur Radio's 6. For more information about CCIR see The ARRL Summa New State of the Art?", Ray Petit, W7GHM, Emission Designators, OST, September 1975, pp. 26-27. "Coher- Handbook (recent edition), Chapter 9, and Demodulation." This is the "promise" of CLOVER. The ent CWThe CW of the Future," Adrian "Modulation Ba Reprinted from the May 1992 issue of CO Magazine, with permission

~~Reprinted with permission, HAL Communications Corporation~~

303 Bill Henry, K9G97, and Ray Petit, W7GHM HAL Communications Corp. Post Office Box 365 Urbana, Illinois 61801 HF RADIO DATA COMMUNICATION CW to CLOVER*

Amateur radio operators have been very efficient code for sending English lan- engaged in data communications guage text because the most frequently used since the first days of spark gap characters are assigned the shortest code transmitters. In fact, the first and ONLY combinations (E, I, S, T, A, N). Morse means of early radio communications re- code is relatively easy to learn, and requires quired ON/OFF keying of the radio trans- only a key and a skilled operator. It is de- mitter using a digital code. That code was, signed for manual operation. of course, the Morse code. With the inven- However, the varying time length of each tion of amplifiers and voice modulators, Morse character and the ON/OFF carrier some amateurs "strayed" into nondigital keying used to send Morse code, make modes (AM, SSB, etc.). However, amateur automatic reception of Morse code by com- radio has seen a resurgence of interest in puters a very difficult task. Automatic com- digital modulation, and there are now many puter Morse code decoders have been de- of us using RTTY, AMTOR, packet, and signed, and Morse receive algorithms im- CW. In this article, we'll discuss these pop- prove with each generation. However, auto- ular digital modes, along with a new one matic Morse decoding still is not equal to called "CLOVER." the decoding skills of a good CW operator. It must also be noted that the abbrevia- Digital modes tion CW in reference to Morse code trans- Morse code mission has helped perpetuate a myth that survives to this day: "Morse code is the Morse code is the original data communi- most bandwidth-efficient mode of com- cations code used by amateur radio opera- munications," and "CW has no bandwidth tors. Often abbreviated as CW (continuous at all." As can be seen in Figure 1, this is by wave), Morse code is transmitted by no means the case! Figure 1 shows the fre- ON/OFF keying of the transmitter carrier. quency spectra generated by a 60-WPM CW Combinations of dots and dashes (short and transmitter using the ARRL-approved rise long key-down times) make up the character and fall times (5 ms). Morse code is ob- codes. Morse code is unique among digital viously not a "zero-bandwidth" emission!' codes in that the length of time required to send each character varies with the charac- RTTY ter sent. For instance, E is one dot, while RTTY (radio teletype), the successor to zero is five dashes. An E is sent in 1/10th Morse code, was originally developed to the time required to send zero. Morse is a automate wire-line message communications. Very complicated teletype machines allowed automation of message handling. 111w article is an update of the paper. "Digital Communications for HF Some of us still use these "mechanical com- Radio-AMTOR! CLOVER." Bill Henry. K90%V-r. and Ray Petit;-;I.. VnGHM. presented to the Amateur Radio Digital Cammuoiatiau puters." As a result-of military require- iaar. St. Louis. Missouri. October 26. 1991. ments during World War II, teletype ma-

~~304 MORSE CODE SPECTRA 0.00000~~

~~ON-MT Keying 01 5 asec Hisetise 25 pits per Second (60 W.P.M.) -12.000~~

~~24.000 W O~~

~~-48.000~~

~~-0.000 -1000.0 X00.00 -200.00 no.00o 600.000 (000.00 FREQUENCY (Hz)~~

Figure 1. Morse code spectra. chines were connected to radio transmitters derv computers and allows unique encoding and'RTTY cache into being. Amateur of 128 different letter, number, punctua- RTTY today still follows the general tech- tion, and control characters. Many of us ex- niques first used in the 1940s (but with no- perimented with ASCII, and soon learned table improvements in terminals and that it's not very robust when used on BE modems).'.' Noise and QRM fehits"often convert a Most amateur RTTY operators use the valid character into a control code that original five-bit feBaudot" (or "Murray") wreaks havoc with our printers (large/small code. The five bit code length limits Baudot type, form-feeds, etc). Most RTTY ama- to a maximum of 32 unique characters-not teurs continue to use the Baudot code. enough to represent 26 letters, 10 numbers, Baudot and ASCII are both "asynchro- and punctuation. This problem is solved by nous" codes. They have START and STOP use of LTRS and FIGS case-shift charac- pulses that allow receive synchronization on ters; each code combination is used twice. each character. Other data modes use codes However, Baudot can't be used to send low- that are "synchronous." They have no er-case letters or an extensive set of control START or STOP pulses. characters. Amateur RTTY can be used at many In 1980, United States amateurs were al- "speeds," but 45 baud (60 WPM) and 75 lowed to use the seven-bit computer ASCII baud (100 WPM) are by far the most pop- code as well.* This code is used in all mo- ular.** When ASCII is used on HF, it's almost always sent at 110 baud (100 WPM). These speeds correspond to a data "through- 'TDae is adore than a We codfuioa as to wbether ASCII is a seven-bit put rate" 6 to 10 or eW-bk code. Wbea waft ASCII as HF at 110 bard, modes aam- of characters-per-second bob seed sevco data bin said ewe "parity" bit. Time are ai& bits bo- (cps). RTTY modes offer no means to cor- eweea the Start and Stop pulses. One4madred and sto baud ASCII uswl- rect transmission errors. y has one surf bit, cgk "dust field bits," said two stop bin, for a total kWh of I I bit units. However, oils Is by m soma "standard." Some HF RTTY is transmitted using Frequency system wed osy sew dsu bitsand no prise trhae the pasky bit is mor- Shift Keying (FSK) of the transmitter radio mally iocased. If co ek46k pessonsa r seas bit 6 to MARE;, the serial frequency. This can be done aynchrooow code will be compedble with coeirisp UARTs an for ckb- by either shift- er aw bk or eipa-bk codes (bit s act m "alwap MARK"). ing a transmitter oscillator directly, or by **For FSK RTTY. AMTOR, and packet radio. the baud rate is equal to I driving an LSB transmitter with audio divided by the time width of ewe data pulse. For erampie, WWPM tones. Most the RTTV base dau palse width of 32 ass. Its baud rate is 1/0.= . 43.45 of modern transceivers that bated, rsualy abbreviated as lS beed. However. CLOVER bb the a+p- include an "FSK" mode really use an inter- bility of sendio8 aam than ow data esnc per dma pulse, by uoap muki- nal audio oscillator to drive LSB transmitter pie phase aid amplitude kvds. Tbetefore. CLOVER hu ore baud rue-31.23 baud per tone parse (123 baud composite)-but several circuits. The standard RTTY "SHIFT" is tbroaabpta rata-18.73 bits-per-se cod (bps) to 730 bp. 170 Hz-the difference between the MARK 305 and SPACE frequencies. Most HF RTTY ters chirp ON and OFF, but data is sent in demodulators use audio tones at 2125 one direction only. The station sending data (MARK) and 2295 Hz (SPACE). RTTY de- is called the ISS (Information Sending Sta- modulators come in many different shapes tion) and the station receiving text the IRS and price ranges. High performance de- (Information Receiving Station). Data al- modulators can be very complicated-and ways flows from the ISS to the IRS. How- expensive.'-' ever, a special OVER command reverses the RTTY has proven to be a convenient and roles of the two stations so data may be popular mode for most of us. In addition to passed in either direction. ARQ mode only its wide use for rag chewing and chasing works in a two-station network. Three or DX, RTTY made a new service possible- more stations may not use ARQ mode with "mailbox" store-and-forward message full error correction. Additional stations handling (MSOs). However because neither may, however, Monitor, but without error Baudot nor ASCII RTTY include error cor- correction. rection, othcr modes like AMTOR, packet, ARQ stations are time synchronized so and CLOVER are more suited for mailbox the transmitting time of one corresponds to use. the receiving time of the other. AMTOR is also a "synchronous" rather than "asynchronous" mode. AMTOR characters do AMTOR not include START and STOP pulses. Accurate AMTOR is an amateur adaptation by timing is important in AMTOR and, for G3PLX of a commercial RTTY mode first this reason, AMTOR controllers must be devised in the late 1950s for ship-to-shore considerably more sophisticated than RTTY use. The commercial version is often called decoder devices. To assure correct timing, TOR (Teleprinting Over Radio) or SITOR the station making the original call is named (Simplex Teleprinting Over Radio). The the MASTER station. It sets the timing for CCIR and FCC call it Direct Printing Ra- both stations for the duration of the ARQ diotelegraph. AMTOR specifications are "link." The station originally called is the defined in CCIR 476-4 and CCIR 625. AM- SLAVE station. TOR adds one mode not described by the AMTOR ARQ mode uses station identifi- CCIR-the "Monitor" or "Listen" mode.'-" er SELCAL characters (SELective CALL). Like RTTY, HF AMTOR uses simple An ARQ mode link is initiated by the MAS- FSK transmitter modulation with 170-Hz TER station that sends the SELCAL of the shift. Commercial SITOR also uses 170-Hz desired station. When a receiving station shift FSK modulation, but with reverse data recognizes its SELCAL code, it responds polarity. AMTOR signals are always sent at with a chirp and the link is established. a data rate of 100 baud. Note that only the designated station will AMTOR digital code has seven bits. The respond, and many stations may Listen on AMTOR code is arranged so all characters the same frequency. A CCIR-476 link re- contain foar MARK and three SPACE data quires only the SELCAL of the destination pulses (called "B" and "Y" pulses in station; a CCIR-625 link exchanges the "AMTOR-speak"). The receiving code con- SELCAL of both stations. The CCIR-625 verter examines each character for this 4/3 SELCAL code is also longer-seven charac- ratio and assumes that the character is in er- ters rather than four for CCIR-476. Both ror if the ratio test fails. This is the error CCIR-476 and CCIR-625 formats are legal detection algorithm of AMTOR. for U.S. amateur use, but most of us con- AMTOR has two primary operating tinue to use CCIR-476 SELCAL codes. A modes: ARQ (Automatic Repeat reQuest) special control END character is used to ter- and FEC (Forward Error Correction). In minate an AMTOR ARQ transmission. ARQ mode, the sending station sends a AMTOR FEC transmissions are one-way burst (or "chirp") of three characters and and may be received and printed by any turns its transmitter OFF. The receiving sta- AMTOR-equipped station monitoring the tion examines each character for the 4/3 bit frequency. A SELCAL isn't used in FEC ratio. If all three characters have the correct mode. Amateurs use FEC mode primarily ratio, the receiving station sends a short for calling CQ, but sometimes also for control signal that means "send next chirp." round-table rag-chewing. If any of the three characters fail the 4/3 ra- FEC mode is much like RTTY in that one tio test, the receiving station sends a differ- station sends his complete message and then ent control code that means "repeat last turns his transmitter OFF to receive the chirp 09 other station(s). FEC mode does not use Note that in ARQ mode, both transmit- three-character chirps or repeat/continue 306 control signals from the receiving station. very slow by modern data standards. If FEC mode provides limited error correction there are no errors to be corrected in ARQ by sending every character twice. The repeat mode, the maximum rate at which data can of each character is spaced in time so a be transmitted is 6.67 characters per second character lost by a noise or interference (cps). This is about 66 WPM, or the equiva- burst may be received correctly at a later lent of 50-baud RTTY. When errors are de- time. The receiving AMTOR controller ex- tected and must be corrected by retransmis- amines the first received character for the sion, this rate slows down. AMTOR, there- 4/3 ratio and prints it when the ratio passes. fore, is not a rapid way to send a lot of If the first transmission of a character fails data. the test, the second transmission of that Finally, due to the unique 4/3 ratio en- character is examined and printed if it is coding of the seven-bit characters, the AM- correct. If both transmissions of a character TOR code supports a maximum of only 36 fail, a blank space (or underline) is printed unique code combinations, including con- to indicate an uncorrectable transmission trol signals. AMTOR uses the same al- error. gorithm as Baudot, and each bit combina- FEC mode isn't as "robust" as ARQ tion is used twice: once in LTRS case and mode, and some errors can't be corrected. again in FIGS case. AMTOR, therefore, However, FEC mode does allow transmis- suffers the same problems as Baudot for sion to more than one other station with transmission of lower-case letters and com- some means of error correction. puter codes. This isn't normally a serious Selective FEC (SEL FEC) is a third mode limitation for ship-to-shore commercial use, used in commercial SITOR. This mode re- or in most amateur applications. However, quires a SELCAL code, much like ARQ it is a big handicap if computer data is to be mode, and will only be received by stations sent.* whose SEL-FEC SELCAL matches that AMTOR, like RTTY and CW, is a fairly sent by the transmitting station. SEL-FEC narrow-bandwidth emission. The measured is commonly used by commercial land sta- spectrum of an AMTOR modulator tions to restrict transmissions to "company (ARQ-1000 and ST-8000) is shown in Figure ships." The ARQ SELCAL and SEL-FEC 2. This spectrum is also representative of SELCAL are usually different sets of cha- that produced by a RTTY station (45 baud racters. To date, most amateurs haven't used is slightly narrower). Due to sidebands and SEL-FEC mode. demodulator filter bandwidths, AMTOR AMTOR and SITOR enjoy widespread stations can be operated with a minimum of use all over the world, and the benefits of 1 kHz spacing between signals if all stations error detection and correction are greatly use 500-Hz wide receiver filters. appreciated. The APLink program developed by WSSMM is widely used to link HF PACTOR AMTOR to VHF packet message networks. As will be discussed later, AMTOR is con- PACTOR is a new development from siderably more reliable for HF data trans- DL6MAA, DF4KV, DLIZAM, and mission than packet radio. DL3FCJ. At this time, only a few PAC- . However, AMTOR is not without its li- TOR units have found their way into the mitations. AMTOR uses an error detection United States, and some features of the new code, but the code is not infallible. In fact, mode are still under development. PAC- a burst error can frequently change the TOR is a modification of AMTOR that states of not only one but two data bits-re- provides moderate speed improvements sulting in the printing of incorrect charac- over AMTOR. PACTOR uses an ASCII ters. This happens infrequently, but it does character set rather than Baudot, and error happen. detection is by means of a CRC (Cyclic Re- Also, the AMTOR throughput speed is dundancy Check) much like that used for AX.25 packet radio. PACTOR also includes an optional data compression mode that can further increase the data speed. OWSWAM and G3PLX are ww aqa with an encoded AMTOR However, the compression algorithm (Huff- code Nut supports upper and Iowa rata hum. man encoding) is language specific; it works "the pears shown in Fkm 2 is a plot of measured data taken tram great on text in some languages, but won't Ow audio output of a HAL ST40W and ARQ-1000 opaaiog in ARQ mode (1704tz sWt. tons - 2125/2295 Hz). The dw hn been oumaic- increase speed on nontext data transmis- aiy shitted to show 0 Hz as the center frequency and 0 ds as the max- sions. Finally, PACTOR includes a speed- imum peak amplitude. The mawuszaw device was a Hewlett-Packard change algorithm in which Modd HP-3561A Druamic Signal Aoayaa. The pmra of a 60-WPM the bit rate of NS baud) FSK RTTY sisal will be simile. bug the MARK and SPACE the data may be increased from 100 baud to pmra will be dishty naemwer- 200 baud when few errors are detected 307 AMTOR SPECTRA 0.00000 MEASURED DATA

~~ICI HAL COMMUNICATIONS AMTOR - ARO MODE~~

~~AR01000WST8000 100 BO. 110 SHIFT -12.000~~

~~-21.000 W O N y~~

~~V -IB.000~~

~~-60.000 I i -1000.0 -600.00 -200.00 200.000 600.000 1000.00 FREQUENCY (Hz)~~

~~Figure 2. Amtor gwxtra.~~

(good conditions). PACTOR is still experi- repeat if the two do not mhtch. This is much mental and promises moderate gains in data the same as.AMTOR ARQ mode and, in throughput on BF radio-two times if 200 fact, packet radio is another form of the baud can be used, and also up to two times generic ARQ class of data transmission when Huffman data compression can be us- modes. Also like AMTOR, packet includes ed. PACTOR, like AMTOR and RTTY, station identifiers, and links only with uses 170-Hz shift FSK modulation. designated stations. However, these iden- As with AMTOR ARQ mode, PACTOR tifiers can be complete callsigns. is an error detection and repeat mode, Unlike AMTOR, packets contain many rather than a true error correction mode bytes or characters-32, 64, and 80 are com- (without requiring repeats). At this time, mon numbers in use-and may have up to PACTOR is an interesting new mode that 255 bytes per packet. Packet radio burst holds good promise for modest im- transmissions aren't evenly spaced in time; provements in HF data transmission over packet transmissions are randomly spaced. At that presently offered by AMTOR. present, packet radio uses CSMA (Carrier Sense, Multiple Access): the controller listens Packet radio and does not transmit if other stations are Packet radio has truly caused a revolu- already sending. CSMA allows many packet tion in amateur radio digital mode opera- radio stations to share the same frequency- tions, thanks to pioneering work by TAPR each responding only to the station with (Tucson Amateur Packet Radio group). which it is linked. This feature was a large Packet radio uses a modification of the boon to early development of VHF packet ANSI X.25 protocol-AX.25.11" radio, but has proven to be a major limita- Like AMTOR, packet data is sent in tion now that thousands of hams are using bursts called "data packets." Unlike AM- the mode. If a lot of stations (10 or more) at- TOR, data is encoded in eight-bit "bytes," tempt to use the same frequency, all traffic and the ASCII code may be sent and received slows and eventually bogs down completely. directly (as well as eight-bit binary data). The On HF, it takes only three or four stations start of a packet includes callsign identifiers to gridlock the channel. and allows specification of repeater paths. Packet radio has become almost the ex- The data packet ends with a CRC check sum clusive data mode used on VHF. Typical number (Cyclic Redundancy Check). The 2-meter VHF FM packet stations use receiving station computes a CRC number 1200-baud, 1000-Hz shift AFSK modems from the packet data it has received, com- based on the Bell 202 modem standard pares that to the CRC sent, and requests a (1200/2200 Hz). A growing number of VHF 308 and UHF packet operations are now switch- line and VHF radio use. However, the very ing to 2400-baud PSK modulation. A few features that make the protocol so useful on UHF packet network relay stations operate VHF, conspire to create big problems for at very high data rates. HF use of AX.25. First, packets are loaded Packet radio has also been attempted on with a lot of "overhead"-non data charac. high frequencies, using 300-baud, 200-Hz ters (callsigns and repeater fields, for exam- shift FSK modulation. HF packet radio ple). Second, packet radio is an error detec- hasn't been a big success, and most of the tion mode, not a direct error correction problems may be traced to the modulation mode. The CRC is computed for the entire format and to the AX.25 protocol itself. packet (including overhead). If the received The modulation format used for HF CRC doesn't match that transmitted, the packet is based on the )3ell 103 300-baud entire packet must be repeated. telephone line modem. While this format is Packet lengths can be set as short as 32 well suited for stable low-noise wire lines characters, but this is very inefficient. Often (and VHF radio), it has serious problems there are more overhead than data charac- for HF use. First, use of simple FSK at a ters in a short packet. We would much rath- 300-baud data rate flies in the face of years er send longer packets (64, 80, or greater) of evidence that ionospheric multipath dis- and improve the efficiency. Sending longer tortion severely affects any modulation in packets increases the time the transmitter is which the base modulation rate exceeds 100 on the air and, more importantly, the time to 150 baud. Multipath time "smearing" over which the CRC is computed, and the often exceeds 3 to 5 milliseconds. Since the time during which completely accurate data width of a 300-baud data pulse is only 3.3 must be received. ms, time smearing irretrievably masks data HF ionosphere disturbances are often of bits. Second, use of the 300 baud rate with a "burst" nature-short, high intensity, a very narrow FSK shift (200 Hz) creates a and widely spaced. One "burst" within a spectrum that is not easily detected. In fact, packet requires the repeat of the complete traditional in-band diversity from separate packet. As the packet length is increased to filters for MARK and SPACE, like those improve efficiency, the probability of dam- used for RTTY and AMTOR, is not pos- age by a burst greatly increases to the state sible. In retrospect, it can be stated that the that no data can be passed. We, therefore, choice of narrow shift FSK and a 300-baud have the contradiction that short packets data rate for BF packet was a poor en- should be used to combat burst interference gineering decision. and noise, but that long packets give better The AX.25 protocol is excellent for wire- efficiency. The ultimate result in both con-

~~HF PACKET SPECTRA 6.0000 MEASLRM DATA W HAL COMMUNICATIONS W PACKET~~

~~PK-232 300 60. 200 SHIFT -12.000~~

~~H - -36.000 aJ~~

~~-.6.000~~

~~0.000 I 1 1 1 1 1 1 1 1 1 1 -1000.0 X00.00 -200.00 200.000 600.000 1000.00 FREQUENCY (Hz)~~

Fipre 3. HF packet spectra. 309 ditions (long or short packets) adds up to quence of smoothly shaped pulses at a sin- very slow transmission of data. HF packet throughput gle carrier frequency. Data is sent in the dif- can be as high as 15 to 20-char- ference acters-per-second in the phase between successive at 300 baud under ideal pulses. conditions. The base data rate of CLOVER-1 is However, what is actually ob- 25 bits served per log, PSK level per second. Ray are throughputs on the order of two uses shaping to four characters-per-second-less of the time pulse to reduce all than CLOVER sidebands AMTOR, which to less than -60 dB operates at 1/3 the baud beyond the rate and in a narrower 100 Hz bandwidth. CLOVER-1 bandwidth! requires relatively simple Further, the "CSMA" concept analog circuits in doesn't the modem, but it makes extreme require- work well on HF. It is easily tricked by ments on the radio equipment. In CLOV- noise bursts and splatter from other HF ER-I, packet the radio frequency stability and tun- signals. Also, CSMA doesn't prevent ing accuracy simultaneous must be maintained within - transmissions by multiple sta- 0.10 Hz! t tions (collisions) Because this is not achievable by due to propagation skip most commercial transmitters zones-the "hidden transmitter or receivers effect" presently available, part of the (that is, your station can't hear CLOVER-1 the interfer- design includes the transceiver itself. ing station but your destination station CLOVER-II can). is a much expanded version of CLOVER-I." HF packet This mode makes heavy radio at 300 baud, 200 Hz use of digital shift is very signal processing (DSP) spectrum inefficient. Figure 3' techniques. CLOVER-II shows the measured bandwidth is ex- spectrum of an HF panded to 500 Hz to better packet modulator (PK-232). match the "nar- The "gentle- row" filters commonly available for men's agreement" is that all HF packet mod- sta- ern commercial transceivers. The CLOVER- tions must be spaced by at least 2 kHz to II "carrier" avoid is a steady sequence of four mutual interference. This interference tone pulses at is caused by the ascending audio frequencies. wide bandwidth of the Each pulse has a duration modulation sidebands of 32 millisec- themselves and the onds; successive pulses reach poor selectivity requirements their peaks at forced on the instants 8 ms apart. The four tone receive modem by the FSK modulation pulses for- are also spaced 125 Hz apart in frequency, mat (300 baud/200 Hz shift). Two AMTOR and stations carefully shaped in amplitude so that can operate without interference in their spectra don't the spectrum required overlap. Data is carried for one HF packet in the difference between station. As we phase and/or shall see, CLOVER allows amplitude of successive four stations pulses at the same to operate in the same band- frequency. These changes width as an HF packet are made only at signal. the instants midway between the In summary, we must comment peaks of that two successive pulses when their amplitudes while packet radio is a wonderful VHF and are zero. As a consequence, the UHF mode, it has very basic usual wide limitations for bandwidth associated with phase modula- HF use. Practically all aspects of today's tion is avoided. The composite signal HF packet signal are wrong is 500 for the condi- Hz wide. The crosstalk between two tions radio operators face daily on HF radio CLOVER circuits. signals spaced 500 Hz apart is less than 50 dB. As with CLOVER-I, varying levels of PSK and ASK modulation CLOVER are used on each tone pulse so data throughputs as high as 750 CLOVER is a new data mode invented bits-per-second are by achieved from a base modulation Ray Petit, W7GHM. Ray's work was in- rate of 31.25 bits-per-log, (level)-per-second. spired, first, by his earlier development and experience with very narrow bandwidth CLOVER-II has a total of eight different coherent CW and then by his observation of modulation formats that may be selected. the many HF packet radio problems we've In order of increasing throughput, these discussed. Rather than modifying existing are: dual diversity BPSK (Binary PSK), modes, the CLOVER design started with a dual diversity FSK, BPSK, QPSK (4-level careful analysis of the unique problems of PSK), 8PSK (8-level PSK), 16PSK (16-level sending data via HF radio. CLOVER is tail- ored to overcome HF radio's unique problems."-:' *Tbe There are two variations Spectra shown in Room 3 is Plot of measured data taken from the of CLOVER: audio output of an AEA PK-232 operating in HF packet mode with 300 CLOVER-I is a 100-Hz baud bandwidth mode; data and a shift of 200 Hz (tones - 2110/2310 He). The data has been numerically CLOVER-II is a 500-Hz bandwidth mode. shifted to sbow 0 Hx as the center frequency and 0 dB the maximum The CLOVER-1 waveform is SS peak amplitude. The measurement device was a Hewkn. a steady se- Pack" Modd HP-3561A Dynamic sigod Analyzer. 310 CLOVER-1I THROUGHPUT

Binary Data Through-put (Bits-pasecond) Modulation 4: BPSK 2:'fSK" BPSK OPSK 8PSK 16PSK SPSK/2ASK Base Rate 31.25 31.25 125.00 250.00 375.00 500.00 MODE 'BDIV" 500.Op 75q K "FDIV" "2F" "4P" 8P "16P' RS I "8P2A" '16P4A- CODE EFF 1 I I I I I I I I I I I I 60% 18.75 18.75 75.00 150.00 225.00 300,00 300.00 75' 75% 23.44 73.44 93.75 18730 450.00 '90" 281.25 375.00 375.00 90% 28.13 28.13 11230 56250 225.00 33730 450.00 450.00 R00" 100% 3113 3113 675A0 125.00 moo 375.00 500.00 500.00 75p.M

COMPARISON HF DATA MODES I- HF Data 7koughput.1 Usable MAXIMUM TYPICAL ERROR ATE OMMON NAME Data Bits Mbps CM'Dps CORRECT P SUMABILITY 45 BD 60 WPM' RTIY 5 bits 6130 None GOOD 75 BD 1 100 WPM RT7Y , Few Etmns ' 5 bits 10150 None 110 BD 100 WPM GOOD, Some Errors ' ' ASCII 7 bits 10170 300 ' None FAIR . Many Elms BD '300 BAUD ' ASCII 7 bits 30210 None VERY BAD , An Ernons 100 BD `AMr0R5rrOR 5 bits 6.67/50 6/30 Yea VERY GOOD, No 300 BD 'HF PACKET' ETmns 8 bits 2WI60 2(16 - 4132 1200 Yes POOR, Many Repeats BD ' V1iF PACKET ' 8 bits 80" 0/0 Yes VERY BAD, No Data No s s Da1 of good or bad data is at ooostsmt rate CP bpsS - BiwPerSecood Data 7koushput Table 1. Tn kai data throughput rata of the trarioas combination of CLOVER modes and codes.

PSK), 8PSK/ASK (8PSK plus 2-level ASK), of the various combinations of and 16PSK/4ASK (16PSK CLOVER plus 4-level modes and codes are shown in Table ASK). As might be 1. expected, the data CLOVER41 also includes self-adapting throughput increases with the complexity of software that can adjust for frequency the modulation, and drift much better iono- and tuning inaccuracies. CLOVER-II sphere conditions are required for the re- com- quires frequency resolution of 10 Hz and plex modes. Typical data throughput rates t will track variations up to t 25 Hz, achiev-

~~CLOVER-II COW SPECTRA WA511RED pATA W HAL CM10AWTIOM ALL MM =a~~

~~-12.000~~

~~_Q.0o0~~

I- , , t nA4 . I -1000.0 -600.00 -200.00 200.000 600.000 1000.00 FREQUENCY (Hz) Figure 4. CLOVER-u spectra. 311 in modern HF transceivers. The interference/distortion can be tolerated. Ef- able most CLOVER-II frequency spectra is shown in ficiency is, of course, inversely related to Figure 4. block length, and the shorter block lengths CLOVER ARQ mode is also automatical- will result in lower data throughput (as in packet). ly adaptive. Ionospheric conditions are measured and the data mode of the sending The number of errors which the Reed- station is adjusted to produce the highest Solomon coder can correct is adjustable. there are three set- throughput possible at the current condi- For each block length, tions. As previously noted, CLOVER has tings corresponding to a maximum of 20, Reed-Solo- 12, and 5 percent of the bytes in the block eight modulation modes, four peon encoder modes and four data block that can be in error without loss of the error-correcting lengths-a total of 128 different and unique blgck. Of course, higher combinations. These combinations provide capacity requires higher coding overhead. the bytes in a block that great freedom for adaptive adjustment in The percentage of bytes efficiency") for swan increments. The CLOVER demodula- are DATA ("coder 75, and 90 per- - tion system dynamically and continuously the choices named are 60, ex- pleasures key received signal-to-noise ratio cent, respectively. When conditions are frequency dispersion, and time dis- ceptionally good and error correction isn't (SIN), correc- persion. Thus, the adaptive protocol can de- required, the Reed-Solomon error termine with a great deal of accuracy which tion algorithm can be completely bypassed- (all parameter should be changed to optimize increasing the efficiency to 100 percent data transmission. bits sent are data bits). unique feature of CLOVER-II is For those who may be wondering how Another 97 the error correction encoding. this complex modulation fits Part of Reed-Solomon and Regulations, let us assure differs notably from that used for AM- FCC Rules This CLOVER modes are indeed in ARQ and packet radio. AMTOR uses you that all TOR conformance with existing rules. CLOVER parity coding (4/3 bit ratio) and packet a station passes only one data stream and is therefore check-sum (CRC) so the receiving The errors and request repeats. not a "multiplex" modulation format. can detect is audio tones Reed- Solomon encoding allows CLOVER modulation output CLOVER's are used to drive the input to an SSB receiving station to fix errors without re- that the This is CCIR mode "J2," quiring repeat transmissions. This greatly transmitter. of CLOVER com- which is allowed. The CLOVER base data increases the efficiency within to AMTOR or packet. Of course, rate is 31.25 bits-per-second-well pared maximum BF limitation. The are linu'ts to the capability of in-code the 300-baud there CCIR emission designator for CLOVER-II error correction and CLOVER will also re- requesting repeats (ARQ-style) when is "500HJ2DEN." vert to restrictions on the error correction system is overloaded. CLOVER places no its alphabet used for sending the data. One of the variable parameters in CLOV- CLOVER accepts any sequence of bytes for block ER-II modulation is the length of the transmission and presents the bytes un- of data sent. This is analogous to packet modified at the receiver. This avoids the length length. However, in this case, block code-specific problems noted for RTTY and and the number of Reed-Solomon correct- AMTOR. blocks able errors are proportional-longer Like AMTOR, CLOVER-II has two pri- can correct more errors without requiring mary modes of operation: ARQ and FEC. packet repeat transmissions. As noted in the Also like AMTOR, CLOVER ARQ mode radio discussion, longer block transmissions uses rigorous timing of the data transmis- lead to higher efficiencies-higher data sions by the two stations. We call CLOVER throughput. Thus CLOVER, unlike packet, ARQ transmissions "twitters." ARQ mode includes an algorithm that is compatible is a two-station link requiring SELCAL ef- with sending long blocks of data at high (full call sign) exchange when linking. CLO- ficiency. Of course, if interference, QRN, VER FEC mode is a "broadcast" format or ionosphere distortion is high, the error receivable by many stations. correction ability of the Reed-Solomon encoding can become "overloaded" and re- Also, if bursts occur peats are necessary. 7be spectra shorn in Flpm 4 is a plot of measured daze taken from the frequently, a long block may contain sever- audio output of a HAL *C=VER-11 Modem, a developmental model. al and also overload the error corrector. The ememble output was centered at 1000 Hz with individual tone center block lengths fre peocies of 812.5.937.5. 10623. and 11173 He. The data bas been Four different Reed-Solomon numerically shifted to sbow 0 Hz as the center frequency and 0 dB as the are available (17, 51, 85, and 255 8-bit ma-um peak amplitude. The mnauaoent devict was a Hewlett-Pack- bytes) so the different conditions of burst ard Model HP-3561A Dynamic Signal Analyzer. 312 As we write this, the CLOVER mode is Like PACTOR, CLOVER is an experi- still under development and working pro- mental mode. Early tests have produced er- totypes are now being tested. To date, on- ror corrected data throughput at 50 cps in the-air results have confirmed the theoreti- typical conditions, and higher levels in good cal work. On typical HF links, CLOVER conditions. To date, CLOVER promises at passes error-free data at rates 10 times least a 10 times increase in data throughput faster than either AMTOR or HF packet. over AMTOR or HF Packet. Average HF data throughput on the order of 200 to 300 bps is realistic. When condi- Error processing tions are good, the throughput can be as Neither CW nor RTTY has any provision high as 500 to 600 bps. for automatic error detection and/or correction. The human mind is a great "interpreter," and often we can "fill in the Mode comparison blanks." This is fine for chit-chat, but is We have discussed a total of six unique totally useless for transmitting nontextual modes that maybe used for transmission of data. HF data (CW, RTTY, AMTOR, PACTOR, AMTOR ARQ mode offers error detec- PACKET, and CLOVER). In the sections tion and correction by means of repeat that follow, we'll compare the performance transmissions. However the. seven-bit ''pari- of these modes. ty , detection scheme isn't infallible, and AMTOR can print incorrect characters. Data throughput AMTOR FEC mode also offers limited error correction by Table 1 shows typical data rates and char- sending each character twice. However, acter throughout that can be expected from if both characters are flaw- the various HF data modes. ed, errors won't be corrected. FEC suffers the same Typical amateur CW speeds are 20 WPM problem with printed undetectable errors as ARQ or slower-a throughput of about 2 charac- mode. PACTOR, ters-per-second. Some operators can send like packet, uses a CRC block error detection and receive code at speeds of 60 WPM and system. Like AMTOR, errors can only higher, but this is the exception. Also, CW be detected and then corrected generally requires manual decoding and by repeat transmissions. The CRC algo- there is no "automatic" error correction. rithm is quite robust and the chances of passing an incorrect RTTY is typically run at 45 baud (60 block are very small. HF Packet also uses WPM), but some MSOs operate at 75 baud a block CRC calculation to detect (100 WPM). At best, RTTY throughput is received errors which then must be fixed by repeat transmissions. 10 characters-per-second. None of -the As RTTY modes offer error noted for PACTOR, the CRC algorithm is correction. very AMTOR, at best, has a throughput of robust. HF packet is, however, very susceptible transmission 6.67 characters-per-second, and slows as er- to errors. This generally results in rors and repeats increase. AMTOR does in- a time-consuming process of sending clude error detection, and most receive er- repeats. rors can be corrected and fixed by repeat CLOVER is the only mode that includes transmissions. AMTOR ARQ throughput error correction in its data encoding. The receiving of 5 cps is probably the "typical" condi- station can correct a limited num- tion. ber of errors without requiring repeat trans- PACTOR offers some improvements missions. If the correctable error limit of the CLOVER mode over AMTOR: up to 13 cps with 200-baud is exceeded, repeat data rate, and even twice that when data transmissions are used in ARQ mode. The Reed-Solomon compression is fully active. However, as algorithm is also very robust and the chances noted earlier, sending data at 200 baud us- for undetected errors is very small. ing simple FSK on HF is risky business and the data compression algorithm will generally not produce a full 2 times speed in- Bandwidth and bandwidth efficiency crease. PACTOR remains to be tested, but Signal bandwidth and its uses are very the "typical" throughput may be on the or- misunderstood parameters of a radio data der 10 of cps. signal. Often, the bandwidth of a mode is As we've noted, HF packet has many specified as that range of frequencies be- problems. As it is now used, the long-time tween the - 3 dB or - 6 dB points on a average data throughput of 20-meter HF spectral plot of the signal. A more realistic packet stations is on the order of 4 charac- measurement of bandwidth is to determine ters-per-second-often, even less. how closely in frequency two signals can be 313 Figure S. CIAVER/AMfOR/packet spectra. placed without mutual interference. Because Figure 1 shows the bandwidth of a CW the signal strengths of two adjacent HF transmitter sending code at 60 WPM. The signals often vary by 30 to 50 dB, and all - 50 dB bandwidth of this signal is ap- data modes (except CLOVER) have wide proximately 800 Hz (t 400 Hz). You can't sidebands, minimum channel spacing is a have two 60-WPM CW stations closer than much higher number than might be in- 800 Hz to each other without potential mu- dicated by the - 3 or - 6 dB spectrum tual interference. This bandwidth scales bandwidth. with speed-a minimum spacing of 267 Hz

~~Figure 6. Minimum spaced HF packet. 314 MINIMUM SPACED AMTOR ... 0.00000~~

~~-10.000~~

~~cell nr' nr' 11,1u11 11,, u11 11r urn 1111 Iq nr' API Ins .r 1,11 uri 1 1 1 1 r 1 1 1~~

~~/ 1 ,~~

~~1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 11 1 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 .~~

~~1 1 1 , 1 1 1 I -4.000~~

~~O1 12 1113 11 !5 M~~

~~-U.000 -.1000.0 3000.0 -1000.0 1000.00 3000.00 5000.00 FREQUENCY (Hz)~~

~~Figure 7. Minimum spaced AMTOP_~~

for 20-WPM signals. Here, the receiver se- Figure 3 shows the measured spectrum of lectivity becomes the controlling point. If an BF packet signal (300 baud, 200 Hz all stations have a 500-Hz receiver filter, the shift). Its - 50 dB bandwidth is greater that minimum signal spacing is at least 500 Hz, 2 kHz! A minimum spacing of 2 kHz is, in regardless of CW speed. Note also that the fact, the practical "500 limit based upon on-the- Hz" receiver filter bandwidth specifi- air experiences. Some BF packet stations cation is almost always the -6 dB band- use a 500-Hz receiver filter. This may help width, not the - 50 dB bandwidth. Auto- reduce interference matic in some cases, but the CW reception is, therefore, probably interference sidebands of an adjacent limited to a minimum HF CW signal spacing of packet signal will still fall within the re- 1 kHz or more. The human brain is adap- ceiver filter's passband. Also, narrowing tive and interpretative, tht and we can indeed receiver bandwidth to 500 Hz may intro- through the QRM." "copy A good CW op- duce data distortion that compounds the erator can often decipher al- CW signals spac- ready poor performance of packet radio on ed as closely as 100 Hz, but only if the two HF. signal strengths are similar. Figure 4 shows the measured Figure bandwidth 2 shows the measured spectrum of of a CLOVER-II signal. CLOVER-II an AMTOR sig- signal. While this is a 170-Hz nals are designed to eliminate sidebands. shift, 100-baud FSK signal, the curve can be The - 3, - 6, and - 50 dB bandwidths of taken as representative of all RTTY and CLOVER are the same-500 AMTOR Hz. The modes (45-baud RTTY will be CLOVER-II demodulator uses receiving slightly narrower, but not by much). Note filters that have passbands identical to the that the - 50 dB bandwidth of this curve is CLOVER-II spectrum. Therefore, approximately 1200 Hz. Practical obser- CLOVER receivers are very resistant to in- vation has shown that, if all stations use terference from nearby signals--M, 500-Hz receiver filters, AMTOR stations RTTY, AMTOR, packet, even another can be spaced as closely as I kHz apart with CLOVER signal. CLOVER signals may be little or no mutual interference. However, if placed exactly "edge-to-edge" either at 500 Hz station in an ARQ link uses his spacing with no mutual interference; 2.4-kHz a SSB filter, the spacing must be in- "guard-band" is not required. Laboratory creased to 1.5 or even 2.0 kHz. The same tests have shown that the co-channel inter- arguments apply RTTY to signals. ference rejection of CLOVER receivers is greater than 50 dB. The data plotted in Fima S. t. 7. and i are the mm dam dun= in Figure 5' shows a combined plot of the Figu .2.3. aced 4. spectrum of AMTOR, HF packet, and 315 Figure E. Minimum spaced CLOVER.

CLOVER signals. The wide sideband nature PACTOR promises modest speed improve- of AMTOR and HF packet is readily appar- ment and ASCII coding over AMTOR. ent, as is the very compact and concentrated CLOVER promises much higher data spectrum of CLOVER. speeds, error correction, and efficient use of Figures 6, 7, and 8 show how much band- our radio spectrum. No one mode com- width is required to support five minimally pletely satisfies all aspects of the others. spaced HF Packet, AMTOR, and CLOVER Amateur radio experimentation with data signals, respectively (see the footnote for modulation techniques is in the midst of a Figure S). Five HF packet signals require 10 new explosion. This renewed interest comes kHz of an HF band, five AMTOR signals as a result of the rapid growth of amateur require 5 kHz, and five CLOVER signals data message systems by VHF packet radio require 2.5 kHz of the band. Put another and the emerging new DSP technology. The way, four CLOVER signals will fit in the ne A and desire to send data more accurate- same bandwidth required for one HF pack- ly and faster via HF radio would not exist et signal, and the data throughput is ten without the growth of the packet radio traf- times higher on each CLOVER signal. fic network. New modulation techniques like CLOVER would not be possible without DSP technology. Although we have Summary concentrated only on the modulation and There are presently four different modu- modem aspects of modern radio data trans- lation and protocol formats used to send mission, it must also be recognized that the amateur data via HF radio: CW, RTTY, next "level"-inter-station communications AMTOR, and packet. Two new modes have protocol-is also undergoing rapid advance- been developed that promise increases in the ments that wouldn't have been possible data speed: PACTOR and CLOVER. The without low-cost and readily available high- six modes differ considerably in perfor- performance personal computers. It can mance, each with its own advantages. CW truly be said that amateur radio is not requires minimum additional equipment to "standing still." Amateur radio continues send and receive (key and good operator); to lead the technology. RTTY equipment is relatively simple and easy to use. AMTOR offers error correction REFERENCES at modest data rates and is very robust for I. Ftgme 1(Mom Code Spectra) is based on measured dm shown to the 1989 ARRL Handbook. Page 9.9. Figure 12. Asociun Radio Achy HF use. HF packet can send ASCII data at league. Newington. Conneaim. when are perfect, but 2. G.W. Henry. K9GWT. "ASCII. Baudot, and the Radio Amateur... good speed conditions W. September 1990. Anmicen Radio Rday league. Newington, Con. quickly falls apart in typical HF conditions. neetiea . pages 11.16.

316 -Understanding 3. Bill Henry. K9GWT, Modern RTTY Codes and Conference Pronedmss. 1990. American Radio Relay League. Newtng Modes," CQ. CQ Publishing. Hicksville, New York, November 1981, too, Connecticut, pages 195-202. 20-25 and December pages 1984, vases 28-32. 23. Ray Petit. W7GHM. "CLOVER is Here." RTTY Journal, Fountain 4. Bill Henry, K9G%VT. "High Tones. Low Tones, Modem Tones ...," Valley, California. January 1991, pages 16-18, February 1991, pages CQ, CQ Publishing, Hicksvil e, New York. November 1982, pages 4248. March 1991, pages 16.17. April 1991, page 10. 5. Bill Henry, K9GWT, "RTTY Demodu4tors". CQ. CQ Publishing, 12-13, Hickr,We NY; November 1983, pages 20.24. 24. Ray Petit. W7GHM. "CLOVER-11: A Technical Overview." ARRL 6. Peter Minister 03PLX, "AMTOR, an Improved Enor-Free RTTY Amateur Radio 10th Computer Networking Colirm" Proceedings. System," QST, American Radio Relay Isaac. Newington, Connecticut. 1991. Amencan Radio Relay League. Newington. Connecticut, pages June 1981, pages 25.27. 125-129. 7. Paul Newland. AD71, "An Introduction to AMTOR,' QST, Ameri- 25. Ray Petit. W7GHM, and Bill Henry. K9GWT, "CLOVER Up- can Radio Rely League. Newington, Connecticut. July 1983. pager Date," RTTYJournal. Fountain Valley, California. January 1992. 11-13. 8. Paul Newland. AD71, "A Lher's Guide to AMTOR," QST. American Radio Relay I ague. Newington. Connection. October 1985. pass 31-34. 9. Paul Newland, AD71. "Algorithms and Methods for SrrOR/AMTOR More on CLOVER Systems," QFX. American Radio Rely League. Newinvon, Ctsomectf- cut. July 1918. pago 9.12. 1 The first CLOVER product available to 10. Bill Henry. K90WT. "New AMTOR Mode." CO. CQ Publishing, Hkkgv k. New York. November 1919. pages 3640. amateurs will be a plug-in card for "IBM- 11. Victor Poor, WSSMM. and Paul Newland. AD71. APMJa4 AMWR compatible" PCs, manufactured by HAL User's Guide, 1989. Available from the ambors. 12. CCIR Recommendation 476.4 (1916), "Daeo-Priming Teleglaph Communications Corp. It will be called Equipment in the Maritime Mobile Service," Racwmmendeiatt and fit- "PC-CLOVER," the "PCI-1000." PC- pores of the CCJR. 1916, Volume VIII.2. Maritime Mobile Swine, Into. national Tdeoom nunicobw union. Geneva. Switzerland. pas ". CLOVER was demonstrated first at the 13. CCIR Recommends loo 615 (1966). "Direct-Priming Telegraph 1992 Dayton HamventionO and targeted Equipment Employing Automatic Identification in the Maritime Mobile delivery thereafter. Patents have Service," Recommendation and Reports of the CCJR. 1916. Volume for soon V1111I-2. Maritime Mobile Swim, Isuenat MM Tdeorsruatmication been applied for CLOVER, CLOVER-1, Union. Gear.-, switudead. Pam 5`99. and CLOVER-11 technology by Ray Petit 14. Hang-Peru Helfert. DL6MAA,and Utrkh Strme. DF4KV. "PAC- and HAL Communications. CLOVER, R-Ra boteiaype with Memory ARQ and Data Caspremon," W. Amctiaa Radio Relay League, Newington. Comemaa, Octobct 1991, CLOVER-I, and CLOVER-II are registered pass 34. trademarks. HAL plans to license CLOVER 13. Martin Clan, DLIZAM. std Peter Made. DL3FCJ "PTC-The PACTOR Controller." W. American Radio Relay League. Newimg- technology as freely and inexpensively as ton. Connecticut. October 1991. pea 7-11. possible to all amateurs and amateur 16.7he ARRL Handbook, Chapter It. omrem aditilm American Radio manufacturers., Relay F ague, Newington, Connection. 17. ARRL Computer Noww"vCotilk emorPtoraedfmgr. Volume I 3'C-CLOVER will include a simple, sin- througb 10. American Radio Relay League, Newington. CemnRrietn. gle-operator application PC program to get IS. Buds Roger. K4ABT. "Packet umes Notebook," smoft column. CQ. CQ Publishing. HickaYk. New York. you started. However, NAL would like to 19. San Horsepa, WAILOU. "Packet Perspective," moothly echmn, encourage all."network" software authors CST. American Radio Relay League, Ncoussion, Claimed st. to write drivers to allow their programs to 20. Bob Ha- N20DL aditor. "Packet signal Ragialer," taoelthly a delta of the Tueaom Ammour Radio Corporation (TAPR). Tucion, itake advantage of the increased perform- Ariroaa. ance of CLOVER. The PCI4000 hardware 21. Ray Pain, W7GHM, -The 'COVERI.FAF ine0 :interface is spmif tally designed to work in HF Data Commmieation Syaem," QMf. Amerso m Radio Relay league. Neirmscm.Connection, July 1990. pas 9.12. Also reprinted .,network and'"windowed" PC environ- in ARRL/CRRL Amemur Radio *it Compueer Newwting Conference ments: An interface protocol definition doc- 1990. Anwian Radio Relay League, Newington. Coaoec- tian, pas 191-194. ument is being prepared for all who are 22. Ray Petit. W7GHM. "FrequencyScable Narrowband Transceiver for Interested. 101003 kHz." ARRL/CRRL Ammew Radio *it Computer Metworkia

~~Reprinted from the Spring 1992 issue of Communications Quarterly, with permission.~~

~~Reprinted with permission, HAL Communications Corporation~~

~~317~~