IAC-09-D1.1.4

FRACTIONATED : THE NEW SPROUT IN DISTRIBUTED SPACE SYSTEMS

J. Guo *, D.C. Maessen †, E. Gill ‡

*Chair of Space Systems Engineering, Faculty of Aerospace Engineering, Delft University of Technology, Kluyverweg 1, 2629 HS Delft, The Netherlands, [email protected] †Chair of Space Systems Engineering, Faculty of Aerospace Engineering, Delft University of Technology, Kluyverweg 1, 2629 HS Delft, The Netherlands, [email protected] ‡Chair of Space Systems Engineering, Faculty of Aerospace Engineering, Delft University of Technology, Kluyverweg 1, 2629 HS Delft, The Netherlands, [email protected]

ABSTRACT This paper provides a survey of current state-of-the-art technologies of fractionated spacecraft, a new architecture for distributed space systems. The survey covers six aspects: architecture, networking, communication, wireless power transfer, distributed computing, and planned missions implementing this architecture. As a result of this survey, the technology readiness level of each of these technology areas is assessed. Finally, research and development activities at the Delft University of Technology in the field of spacecraft formation flying are introduced. Focus is given to areas where fractionated spacecraft share technologies with formation flying spacecraft.

eXchange) [5]. Eventually in 2008, the DARPA INTRODUCTION System F6 contracts were issued to teams headed by The Boeing, Lockheed Martin, Fractionated spacecraft represent a novel Northrop Grumman and Orbital Sciences, for the architecture for distributed space systems. Unlike preliminary development [6]. constellations or formations, where similar This paper attempts to provide a comprehensive spacecraft are spatially distributed, a fractionated overview on current state-of-the-art technologies spacecraft distributes the functional capabilities related to fractionated spacecraft, including the of a conventional monolithic spacecraft amongst work being performed at the Delft University of multiple heterogeneous modules which perform Technology (TU Delft), The Netherlands. The distinct functions and interact through wireless paper consists of two main parts. The fist part is communication links. Although doubts on its a survey of up-to-date research on fractionated economics exist, the fractionated spacecraft is spacecraft. As a truly networked system of attracting more and more attention from systems in space, the fractionated spacecraft academia, industry and governments due to its faces a lot of challenges, such as wireless power advantages of rapid response, enhanced mission transfer, autonomous self-forming networks, and and in-orbit robustness, potential for mass so on. Recent advances in these technologies are production, flexibility with later added features reviewed, with the focus on their availability for and lowered mission recovery costs [1]. In some fractionated spacecraft. Special focus is put on sense the fractionated spacecraft may be ongoing space missions related to fractionated regarded as a game-changing event in the spacecraft. The second part of this paper briefly history of space systems, just like the introduces research and development activities at revolutionized data communications. TU Delft with a focus on distributed space The term “fractionated spacecraft" is relatively systems. As an emerging branch of distributed new and has been coined in 2006 by Owen space systems, fractionated spacecraft can be Brown and Paul Eremenko in a series of papers regarded as sharing some technologies with [1, 2, 3]. In August 2006, a workshop dedicated formation flying, such as wireless communication, to fractionated spacecraft was held by Defense networking and distributed computing. Potential Advanced Research Projects Agency (DARPA) at synergies for technology and programmatic Colorado Springs [4]. In 2007, DARPA announced extensions of TU Delft's work in these areas are soliciting proposals for a program entitled System given. F6 (Future Fast, Flexible, Free-Flying, Fractionated Spacecraft united by Information

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 1 SURVEY ON FRACTIONATED SPACECRAFT fractionation, which allows reality-based focus on specific engineering. Mission-specific functional The generalized concept of a fractionated elements are divided into appropriately sized spacecraft is to break a large monolithic fractionated blocks, mission-support common- spacecraft into smaller homogeneous or function elements are distributed across multiple heterogeneous modules [7]. For homogeneous nodes, and the detailed mass properties of each fractionation, the large spacecraft will be element within the nodes are compared with the replaced with a cluster of identical small known Boeing 601HP mass budget. Rooney spacecraft which can function independently from concluded that the efficiency of fractionation is each other (e.g. formation flying ). For very dependent on the type of the mission and, heterogeneous fractionation, the large spacecraft therefore, geostationary communication missions will be re-constituted by a cluster of wirelessly are not appropriate for fractionation from the interconnected spacecraft modules that each has cost standpoint. a unique function such as Guidance, Navigation In a series of papers presented by Mathieu and and Control (GNC), and command and data Weigel [7, 11, 12], another quantitative way for handling. These heterogeneous modules fly the assessment of fractionated space freely in approximately the same orbit. In most architectures is provided, using Multi-attribute references, including this paper, “fractionated Trade-space Exploration. This method is a spacecraft” only refers to the latter definition. customer-centric approach because it assesses Although the fractionated spacecraft itself is a fractionated architectures in terms of the revolutionary concept, it is still built on past attributes. For the purpose of valuing the achievements. From the other side, the attributes, a set of possible customers are fractionated spacecraft also urged the emerging interviewed for an imaging, a communication and of new technologies. Owen identified four key a navigation mission, followed by a multi- technologies for fractionated spacecraft: attribute utility analysis. Non-traditional networking, wireless communication, cluster attributes, including maintainability, scalability, flight, and distributed computing [8]. In this flexibility and responsiveness are incorporated. section, a review of up-to-date work on The architectures are described by a set of fractionated spacecraft is conducted from six design parameters at both spacecraft and aspects: architecture, networking, wireless fractionation levels, which forms the design communication, wireless power transfer, vector. Through varying the values of the design distributed computing, and planned missions. parameters, e.g. numbers of infrastructure modules, or type of power fractionation, the Architecture trade spaces of the architectures are explored. An important conclusion of this research is that Research on the architecture of fractionated customers would choose fractionated spacecraft spacecraft so far focused on the assessments of if non-traditional attributes are valued enough, the architecture. The first article about this can which corroborates the development of be dated back to 1984 by Molette et al. [9]. In fractionated spacecraft. that paper, two main space segment concepts were assessed: the cluster of cooperative Brown and his collaborators investigated the satellites and a large platform assembled in orbit. architecture of fractionated spacecraft both in The first one is similar with the concept of qualitative and quantitative ways. In [1, 3], he fractionated spacecraft. Different concepts are provided a detailed qualitative discussion about characterized by the required number of modules the advantages and disadvantages of to be launched, the type of launcher, and the fractionated spacecraft, and claimed that the new subsystem to be developed. The main flexibility and the responsiveness of a conclusion of this research is that the cluster of fractionated system will exceed any penalties. In satellites is much less cost effective than the [2], Brown et al. introduced the value model, assembled platform. However, this conclusion is which associates the benefit (value) delivered to not convincing because the attributes such as the user with a given fractionated system adaptability, flexibility and growth potential are architecture. In [13], a methodology to calculate not counted in a quantitative way in their cost- the total system lifecycle cost under uncertainty based econometric framework. was presented. The philosophy of this method is: the system lifecycle cost is a random variable Rooney performed a quantitative assessment of with a probability distribution caused by spacecraft fractionation using Design Rules [10]. uncertainties throughout the lifetime, and this In his work, a known-state spacecraft, i.e. a random variable is the basis for comparison Boeing 601HP geostationary communication between different architectures. According to this , is used as the example to evaluate methodology, the lifecycle costs of alternative

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 2 fractionated architectures were derived using • Networking and multiple access Monte Carlo simulation incorporating launch and capability; component failure uncertainties. It was found • Reliability; that the total lifecycle costs of fractionated • Cost effective; spacecraft are comparable to monolithic spacecraft if the individual modules’ design • Unsymmetrical forward and return links lifetimes are longer. Based on the work in the (different data rates); value-based technology, Brown et al. introduced • Limited bandwidth; the value-centric design into fractionated • Variable return trip times (RTTs) due to spacecraft [14, 15]. The basic idea of the value- changing inter-satellite distances; centric design is to formulate the design of • fractionated spacecraft as an optimization Single event upsets; problem. The fractionated architecture can be • Antenna obscuration and; defined by a relatively small set of tradable • Limited (processing) power, program architectural design parameters such as “degree memory, and data buffering. of fractionation” and “distribution of nodes across According to these criterions, some interesting modules”, and a set of attributes, consisting of conclusions are drawn: ATM does not adequately cost, flexibility and robustness, is folded into the support multiple access, TCP/IP and CCSDS SCPS objective function. Then by given different sets of are too high up the protocol stack to be design parameters, the objective function can be considered as a lower layer protocol, CCSDS AOS evaluated either using the system lifecycle costs is not intended for space-space links, and the method in [13], or using the direct valuation of IEEE 802.11 physical layer would need to be flexibility and robustness. entirely revamped to meet physical layer Networking requirements. The remaining options are the HDLC-based X.25/LAP-B and CCSDS Proximity-1. Like connected computers on the ground, each CCSDS Proximity-1 has many advantages over module of a fractionated spacecraft can be X.25/LAP-B since it was specifically designed for regarded as a node, and the whole fractionated space-space links. However, when Kusza and spacecraft forms a network connected by Radio Paluszek published their paper, CCSDS was still in Frequency (RF) or laser links. If needed, the Red Book stage (currently it is Blue Book!!) and elements of the ground segment can also be HDLC had decades of commercial production and treated as network nodes. existing engineering expertise, even though it Brown pointed out that the fractionated was not intended for ISLs. An experimental spacecraft should be an autonomous, self- comparison of two similar implemented systems forming network of nodes [1], which means that including a cost estimate was needed to if new modules or new spacecraft join the determine whether the benefits built into CCSDS network, they can be integrated in a “plug-and- Proximity-1 will outweigh the availability of play”-like fashion just as for a ground based existing standards such as HDLC or IEEE 802.11. . Some existing technologies, A Spread Spectrum (SS) link (e.g., Direct such as Internet Protocols (IP), could possibly Sequence, Frequency Hopping) is desired for realize this in space. However, so far the work multiple access, security (jamming, primarily focuses on formation flying or satellite- eavesdropping), resistance to degrading and ground networking rather than fractionated multipath effects, and to allow inter-satellite spacecraft. ranging. If a widely supported protocol is used (e.g. CCSDS Prox-1), communication with Kusza and Paluszek [16] compared lower layer protocols for Inter-Satellite Links (ISLs) in an different spacecraft close to the ‘own’ formation becomes an option. attempt to identify a protocol for widespread ISL use such that: Another interesting and also very exciting work • Interaction among different has been reported by Hogie et al. [17]. In that paper, issues related to the use of standard IP constellations is possible and; for satellite communication are discussed, • The addition of new satellites to existing including: constellations is simpler. • Lower layer protocols that deliver data The protocols considered are X.25/LAP-B, onboard the spacecraft and over the TCP/IP, ATM, IEEE 802.11, and CCSDS Proximity- space link; 1, SCPS, and AOS. The criterions are • comprehensive, including: Network protocols that provide global addressing and data routing among systems;

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 3 • Transport protocols to support end-to- communication itself is already a very broad field, end delivery, and; this subsection will only discuss works which • Application protocols to support have been or will shortly be demonstrated in operational needs. space. In the late 1990s, a reference system Intra-satellite wireless communication is an architecture for the space and ground segments approach to reduce the amount of wiring and to of future IP missions was developed within the simplify the integration and test. The first intra- OMNI (Operating Missions as Nodes on the satellite communication was realized on space Internet) project. The key to this whole shuttle missions STS-83 and STS-94, both in architecture is that it is built upon the protocol 1997 [21]. On those mission, 3 Remote Sensor layering concepts of the ISO OSI network Units (RSUs) were installed on longeron rails in reference model. Based on this research, in the payload bay and formed a dynamically February 2000 a standard IP stack was ported to reconfigurable network. RF tests showed good the UoSAT-12 spacecraft developed by Surrey reception throughout the flight-deck and the mid- Satellite Technology LTD. (SSTL), England. In deck. Later, on the International Space Station April 2000 a series of tests, such as on-orbit (ISS), a Wireless Instrumentation System (WIS) testing of UDP telemetry delivery, NTP (operating was installed, which has a 200mW WLAN module over UDP) and FTP (operating over TCP), were working at 900MHz with a data rate of 2Mbit/sec. successfully completed with UoSAT-12 [17]. Next to these examples, a few wireless instruments such as Micro-Strain Gauge Units A further step on space networking beyond (MicroSGUs) and Micro-Triaxial Accelerometer UoSAT-12 was seen from Wood et al. [18]. Units (MicroTAUs) were also flown on space Within the CLEO (Cisco router in Low Earth Orbit) shuttles. project, a commercial Internet router launched into space in September 2003 onboard the UK- Although intra-satellite communication tests on DMC satellite. Until 2008, CLEO has been in orbit space shuttles and the ISS have been successful, for over five years and has been tested in orbit so far this technology has not been used on a for over four years. CLEO has been powered up satellite. The only exception is a very small 3 for use more than one hundred times. The IPv6 satellite, Delfi-C , which was launched in April 3 and IPSec have been tested successfully onboard 2008. Delfi-C is equipped with two autonomous UK-DMC in space. This shows that these wireless sun sensors, each measuring additions to the Internet protocol, developed for approximately 60mmx40mmx20mm and having a terrestrial use, can also be used successfully half-sized GaAs solar cell as its own local power onboard satellites [18], and possibly on supply, making them independent of the fractionated spacecraft. satellite’s electrical power system [22]. The intra- satellite communication is achieved through a The work on networking is rarely directly related wireless Radio Frequency (RF) link. This link has to the fractionated spacecraft. Ivancic [19] and been made using a modified COTS transceiver Wood [20] both discussed the possibility to use operating at 91.5MHz, which enables modularity existing COTS Internet technology for networking and flexibility, reduces system mass and volume the sub-parts of the fractionated spacecraft. This due to the absence of cables and connectors and “mobile-IP” technology has many advantages is an enabler for “plug and play” construction of such as low cost, a large professional knowledge satellites. The in orbit experiment demonstrated pool, and is extensively tested on a daily basis. the feasibility of the wireless link (immunity to One or more (for security) VMOCs (Virtual disturbances; no interference with other Mission Operations Center) can be used to allow equipment) and the operation of the sun sensor system operators and users to be remote. The under variable power supply. fractionated spacecraft can use routers to form a LAN (Local Area Network) between its own sub- Regarding inter-satellite communications, the parts while communication with the ground most famous examples are NASA’s Tracking and station(s) is performed using WAN (Wide Area Data Relay Satellite System (TDRSS), operational Network). DTN (Delay Tolerant Networking) is since the 1980s, and the Iridium global needed when e.g. parts of large image files are commercial mobile communication system, downloaded to different ground stations. launched in 1997 [8]. TDRSS enables spacecraft in any orbit to transfer data forward or backward Wireless Communication to other spacecraft or the ground. Iridium allows users on the ground or in the air to communicate The wireless communication for fractionated with others anywhere on the planet using only a spacecraft is an extension and synthesis of the pocket-sized mobile phone. Next to RF technologies of both intra-satellite and inter- communication, the optical communication, satellite communications. As the space wireless

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 4 especially laser communication, is very promising auxiliary power mode for clusters due to its relatively low power consumption. or fractionated spacecraft. Inter-satellite laser communication technology for Jamnejad and Silva also investigated microwave data relay has been successfully tested in space power beaming strategies for fractionated (see Tab. 1) and will be ready soon on all spacecraft [29]. Some scenarios have been planned ESA data relay satellites. considered, and various issues, including antenna Tab. 1 Space Optical Communication Tests aperture sizes, separation distances, frequencies Test Transmitting Receiving Data and power levels have been discussed. According Date Satellite Satellite Rate to the authors, beam shaping can be used to 2003 SPOT-4 Artemis 60Mbps optimize the transmission efficiency while minimizing the interference and power spillover 2006 OICETS Artemis 60Mbps through sidelobes. The utilization of defocusing 2008 TerraSAR-X NFIRE 5.6Gbps phase arrays, reflector systems, and reflect arrays is proposed. The authors also suggested Wireless Power Transfer several technology investment areas, including efficient rectenna diodes, phased array antennas, Although the physics are identical, wireless high breakdown/low loss transmit filters, phase power transfer is slightly different from wireless shifters, and retrodirective systems. communication: for the prior one, wireless power transfer efficiency is the most important, while Under a DARPA contract, Turner at Space for the latter one the signal-to-noise ratio is Systems/Loral studied a new concept for focused on. Compared with wireless transferring power within a fractionated system communication, the work on wireless power [30]. His idea is to use a “mother ship” or transfer is still in the infant stage. Even on the Resource Vehicle (RV) with deployable solar ground, wireless power transfer over distances arrays to collect solar energy centrally, and then comparable to a few times the diameter of the distribute high-intensity beams of unconverted device has just been realized [23]. For concentrated sunlight to high-temperature fractionated spacecraft, a far-field power transfer compact receivers on “daughter satellites” or is required to achieve longer distance, often more Mission Vehicles (MVs) wirelessly. The MVs use than 100m due to safety consideration. heat engines to generate power from heat stored Therefore, transferring power in space still has a in receiver reservoirs, providing routine and long way to go. contingency energy storage. According to Turner, the power transfer efficiency of this Power transfer via RF, typically in the microwave “concentrated light distribution” scheme is range, can be made more directional, allowing around 25%, much higher than those of longer distance power beaming. A rectenna may microwave (3%) or active optical (4%) schemes. be used to convert the microwave energy back Therefore, this technology is very promising for into electricity with more than 95% conversion wireless power transfer in fractionated efficiency [24]. Power beaming using microwaves spacecraft. has been proposed for the transmission of energy from orbiting solar power satellites to Earth [25], Distributed Computing and the beaming of power to interplanetary spacecraft has also been considered [26]. A distributed computing system may be defined Lafleur and Saleh performed a system-level as one in which hardware or software feasibility assessment of microwave power components located at (geographically) different beaming for small satellites within the computers communicate and coordinate their fractionated spacecraft context [27, 28]. They actions through a wired or wireless network [31]. utilized parametric models of spacecraft power as Distributed computing onboard a single a function of 11 design variables, and then found unmanned spacecraft has been successfully feasible designs through Monte Carlo design demonstrated in space. The first ever spacecraft trade-space sweeps. Several optimistic utilizing distributed computing were NASA’s assumptions were made, including only two Voyager 1/2, both launched in 1977, which make spacecraft or modules in the system, and no use of three RCA 1802 CPUs running at 6.4 MHz pointing losses. Despite these assumptions, it has [32]. Each concentrates on processing different been found that the design space for power functions, such as attitude control, data beaming is severely constrained. Feasible designs formatting, and commanding. Another NASA still require high transmission frequencies, large spacecraft, the Galileo probe, which was launch antenna diameters, and stringent proximity in 1989, has dual processors for attitude control distance. Therefore, Lafleur and Saleh suggested and six RCA 1802 Cosmac microprocessor CPUs to only consider microwave power beaming as an (four on the spun side and two on the despun

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 5 side; each CPU was clocked at about 1.6 MHz, the agents and the nodes. Each node has and fabricated on sapphire) in a network for predictable resources and services, e.g. compute, command and data handling [33]. Both storage, communication, sensing, energy. The spacecraft were designed for long-duration, following problem areas are discussed: autonomous flight, a goal difficult to attain 1. Resource allocation : how to schedule without the use of distributed computing. tasks based on current requirements Townsend et al. [34] and Palmintier et al. [35, and resources? A decision algorithm is 36] designed, implemented and tested a needed. distributed computing architecture for the 2. Groups and membership : How to Emerald nanosatellite formation flying mission. autonomously define an application and The system is composed of a network of COTS how to identify the pieces of an PICmicro® processors linked by a hybrid bus application in a changing environment consisting of a high-bandwidth Phillips I 2C data (spacecraft degrade/are removed/are bus and a low-bandwidth Dallas Semiconductor added)? “1-wire” microLAN. According to the authors, this 3. Communication : How to communicate system significantly improved the development between agents in a secure and reliable work of the Emerald satellite. However, it seems manner? that so far this system only works on a single 4. Failure tolerance : Fault Emerald satellite, although the authors claimed detection/prediction, containment, and that it could also benefit multi-satellite flight recovery are needed as well as dynamic operations. resource allocation. Fault recovery can Distributed computing onboard a cluster of also mean that a replacement spacecraft spacecraft has not been realized so far. This is is added, preferably as quickly as partially due to the complexity of inter-satellite possible. Failure prediction helps to do wireless communication and networking, and that. partially because of the fact that the spacecraft 5. Security : Isolation between nodes and cluster itself is also a relatively new concept. agents is needed to prevent the However, limited work has already been done in spreading of malicious software or bugs. the context of formation flying and, even for The system must be resistant to fractionated spacecraft. spoofing, eavesdropping, denial of service, and rogue nodes. Jallad and Vladimirova [37] proposed the 6. Programming model and tools: A deployment of distributed computing on close programming language, API (Application formation flying missions. Two distributed Programming Interface), and computing paradigms, named Client-Server and development tools need to be mobile agent, are compared, and eventually the developed/selected. A test framework, mobile agent paradigm is selected for the simulators, and model checkers are also application. Meanwhile, in order to evaluate and needed. compare distributed computing paradigms, a 7. Standardization and interoperation : real-time testbed was developed, which utilizes Expect different satellites from different LEON-3 processors, the open source SPARC v9 vendors to be added over time. To allow architecture CPU core, and the Real-Time for smooth interoperation, standards Executive Multi-processor System (RTEMS) need to be defined and rigorously operation system. The main contribution of this implemented. Also, interoperability tests work is the creation of middleware which is an (testing and formal checks) need to be add-on to the existing Real-Time Operating performed. System (RTOS) and allows tighter coupling All these problems need to be solved before such between the nodes in the distributed system. a scheme can be implemented on fractionated Golding and Wong [38] investigated adaptive spacecraft. distributed computing for fractionated space systems. They regard adaptive distributed Fractionated spacecraft Mission computing as multiple distributed applications As the fractionated spacecraft is such a new that all use resources at multiple nodes concept, few missions have been proposed based (spacecraft). The applications have to on this revolutionary architecture. So far the only automatically adapt themselves as needs and fractionated spacecraft mission that has been resources change. An agent-based architecture is announced is Pleiades [39]. considered for each application. Each agent has a different role and the agents are controlled by a The Pleiades system consists of five 225-kg manager. Middleware provides the link between modules (named Atlas-class modules) and two

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 6 75-kg modules (named Pleione-class modules), robustness, and mass production effects will i.e. a total of seven modules. As shown in Tab. 2, exceed any penalties. the completed Pleiades system will accommodate In the field of space networking, exciting three Precision Pointing Payloads (PPPs) and a progresses have been achieved recently in space. single Coarse Pointing Payload (CPP). These The existing COTS Internet technology developed payloads will share two TDRSS transceivers, two for terrestrial use, such as commercial Internet Solid State Recorders (SSRs), two high router and the standard IP, has been successfully bandwidth downlinks, and two high performance demonstrated onboard satellites. By given a Mission Data Processors (MDPs). The whole reliable ISL, a very promising solution for system will be launched into orbit in three fractionated spacecraft networking is to adopt launches. After each launch, the in-orbit modules “mobile-IP” technology in space, and use will have some capabilities to implement the LAN/WAN for inter-module and satellite-ground mission, and after three launches the system will communications. The DTN could also play a key be fully available. Another advantage of using role for networking between heterogeneous multiple launches is that the most mature sensor elements due to its capability of overcoming can be launched first to allow time for the network disconnection. As these technologies development of the less mature sensors. have been extensively tested on the ground on a Tab. 2 Pleiades Modules [ 39 ]. daily basis, they are believed to be ready for Launch 1 2 3 space usage soon. Module A reliable space wireless communication link is the precondition of realizing other enabling technologies. RF-based inter-satellite

e Merope Electra Electra Maia Celaeno Taygete Taygete Alcyone Alcyone Asterop communications have already been widely used Mass[kg] 225 225 225 75 75 225 225 by TDRSS and Iridium, and the inter-satellite PPP Y Y Y laser communication also has been tested in orbit CPP Y and will be available for applications within 4 years. The intra-satellite communication is less TDRSS Y Y mature, and only has been demonstrated in orbit SSR Y Y recently. However, similar with the networking, High BW Y Y the extensive terrestrial applications of the Downlink technology of wireless communication will speed MDP Y Y up the procedure of space applications. At the time LoBosco et al. published their paper, Wireless power transfer is the biggest challenge Pleiades was still halfway in the preliminary among the four enabling technologies. It has just design. been demonstrated in special environment for terrestrial near-field applications, and the space Summary of the Survey far-field transfer, which is necessary for So far the up-to-date developments in the field of fractionated spacecraft, is still in the stage of fractionated spacecraft have been reviewed, exploring feasible technical concepts. Although which covers the architecture, the planned some research indicates that transferring power missions, and the four enabling technologies: via RF is not the best choice for fractionated networking, wireless communication, wireless spacecraft, it is still the mainstream and can be power transfer and distributed computing. improved by using a rectenna to covert the microwave energy back into electricity. Another In the research of fractionated spacecraft possible way is to adopt the “concentrated light architecture, the value-based methodology has distribution” scheme for higher power transfer already been the mainstream for architecture efficiency. However, all these approaches need to assessments. There were indeed some doubts be further investigated and tested on the ground about the fractionation, however these negative before the in-orbit demonstration. opinions are more based on qualitative assessments and visibly struggle with their The distributed computing was successfully inability to quantify non-traditional attributes utilized onboard a single spacecraft many years such as adaptability and flexibility, which actually ago. Its application on a cluster of spacecraft is are the most important advantages of the restrained by the complexity of inter-satellite fractionated spacecraft. According to the results communication and networking. The agent-based of value-based assessments, if these non- architectures are investigated in the context of traditional attributes are considered properly, the fractionated spacecraft. As the agent technology benefits brought by flexibility, improved has been demonstrated onboard Deep Space 1, it

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 7 can be a very promising solution for distributed computing within a fractionated spacecraft. Regarding planned missions, so far there is only one demonstration mission announced for fractionated spacecraft. It will be a milestone of the space history; however the details about the utilization of the four enabling technologies in this mission are still not very clear. Especially the technological challenge brought by wireless power transfer is not well addressed, which may be a bottleneck of the whole mission. Fig. 1 Delfi-C3 Nanosatellite . Based on the above statements, the results of After launched at 28 April 2008 from Sriharikota 3 the survey are partially summarized in Tab. 3, launching site in India, Delfi-C successfully which only covers the four enabling technologies. implemented many experiments and is still This assessment was performed according to ESA operational when this paper is published. An 3 and NASA’s definitions on Technology Readiness important experiment onboard Delfi-C is the first Level (TRL). ever in-orbit demonstration of autonomous Tab. 3 Readiness of Enabling Technologies. wireless sun sensors. As introduced earlier in this paper, there are two autonomous wireless sun Technology TRL * Explanation sensors installed on two opposite sides of Delfi- Networking 5 Partially Demonstrated C3. The sun sensors have built-in RF module, Wireless 5 Partially Demonstrated which is based a COTS transceiver and can Communication transfer data to the satellite’s Command and Wireless Power 2 Even not used on ground Data Handling System (CDHS) wirelessly. The Transfer success of the demonstration proves that intra- Distributed 4 Mature for ground satellite wireless communication can be realized Computing applications not only on big spacecraft, but also on such a * TRL: Technology Readiness Level (ESA&NASA) small satellite. After adaption, this technology could also be used for the inter-module wireless communication of fractionated spacecraft. SATELLITE R&D ACTIVITIES AT TU DELFT FAST Satellite research and development activities at TU Delft are centered around the chair of Space FAST (Formation for Atmospheric Science and Systems Engineering (SSE). The group is Technology demonstration) is a cooperative particularly focused on engineering of micro- and Dutch-Chinese formation flying mission led by TU nano-satellites and distributed space systems. Delft in the Netherlands and Tsinghua University This comprises intra-satellite wireless in China [40]. It is expected to be the first communication, formation flying missions, and international micro-satellite formation flying space wireless sensor networks. This section mission to achieve objectives in three distinct introduces selected aspects which may be fields: technology demonstration, earth science considered relevant for fractionated spacecraft. and space education. As shown in Fig. 2, in FAST mission, two spacecraft, i.e. FAST-D (being Delfi-C3 developed in Delft) and FAST-T (under

3 development in Beijing), will demonstrate Delfi-C is the first Dutch university satellite, Autonomous Formation Flying (AFF) using ISL which was developed by a group of students in and, furthermore, perform earth observations SSE and other chairs in TU Delft with the with more scientific returns due to formation supervisions from staffs [22]. As shown in Fig. 1, flying. the satellite is la cube with the size of only 100mmX100mmX300mm and the mass of 2.2kg.

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 8 challenges to inter-satellite communication, distributed computing, and self-configurable networking. Another example is the QB50, an international network of 50 CubeSats for multi-point, in-situ measurements in the lower thermosphere and re- entry research [44]. All 50 CubeSats will be launched together on a single launch vehicle into a circular orbit at about 300km altitude, inclination 79º. Although this project aims to use 2unit CubeSats, some technologies such as inter- Fig. 2 The FAST Mission . satellite communication and networking still could For the FAST mission, the inter-satellite be demonstrated. communication will be realized through a RF- based direct ISL operating at S-band. Meanwhile, CONCLUSIONS space-based distributed computing technique will also be demonstrated by the FAST mission to As a new sprout in the field of distributed space explore the possibility of fully utilizing the systems, the concept of fractionated spacecraft is computing powers of AFF small satellites for attracting more and more attention. This paper onboard processing. The concept of space-based provides a survey of state-of-the-art technologies distributed computing will utilize the processors of fractionated spacecraft which covers six on each satellite as a true distributed computer. aspects: architecture, networking, wireless For the FAST mission this indicates a space- communication, wireless power transfer, based computer with (at least) two distributed computing, and the planned missions. geographically distributed processors. The types The results of the survey, i.e. Tab. 3, of problems under consideration for such a Tab. 3Tab. 3show that some of the enabling distributed computing network are typically large, technologies have either been successfully dense linear algebra problems. This is because demonstrated in orbit, or already achieved high these operations can be processed using blocked technical readiness. It is also found that the algorithms, which are well suited for distributed biggest technical bottleneck for fractionated and, possibly, heterogeneous systems. An spacecraft is the wireless power transfer. important experiment will be the real-time orbit This paper also briefly introduces the relevant determination of the FAST formation, because research and development activities at TU Delft. the size of this problem could be very large (i.e. The successful in-orbit demonstration of intra- thousands of parameters using thousands of satellite wireless communication through the observations) and beyond the capability of a Delfi-C3 nanosatellite, and the planned single processor [41]. Details about the FAST autonomous formation flying, inter-satellite mission can be found in [42]. communication, distributed computing Under the framework of the FAST mission, demonstration through the FAST formation flying several aspects of formation flying, such as inter- mission as well as other activities can benefit the satellite communication, distributed computing, development of fractionated spacecraft. cluster flight, and reconfigurable networking, will be validated, which will also significantly REFERENCES contribute to fractionated spacecraft. 1. Brown, O. and Eremenko, P., Fractionated Other Activities Space Architectures: A Vision for Except the above two projects, SSE is also Responsive Space , Proc. of 4th involved in other activities related to distributed Responsive Space Conference , AIAA-RS4- space systems. 2006-1002, Los Angeles, CA, 2006 One example is the Orbiting Low-Frequency 2. Brown, O., et al., Cost-Benefit Analysis of a Antennas for Radio Astronomy (OLFAR) project, Notional Fractionated SATCOM which presents a new concept of a low frequency Architecture , Proc. of the 24th AIAA radio telescope in space using a cluster of small International satellites with the virtual aperture of Systems Conference, AIAA-2006-5328, approximately 100km [43]. As the data San Diego, CA, June 11 – 14, 2006 correlation must be done in space, distributed 3. Brown, O. and Eremenko, P., The Value processing with centralized downlink transmission Proposition for Fractionated Space is the preferable option, which will bring Architectures , Proc. of AIAA Space 2006

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 9 Conference & Exposition, AIAA-2006- 15. Brown, O. and Eremenko, P., Application of 7506, San Jose, CA, September 19-21, Value-Centric Design to Space 2006 Architectures: The Case of Fractionated 4. Brown O., Fractionated Spacecraft Spacecraft , Proc. of AIAA SPACE 2008 Workshop: Visions and Objectives , DARPA Conference & Exposition, AIAA-2008- Fractionated Spacecraft Workshop, 7869, San Diego, CA, September 9-11, Colorado Springs, August 3-4, 2006 2008 5. “System F6 ”, Broad Agency Announcement, 16. Kusza K.L. and Paluszek M.A., Intersatellite DARPA, 2007. Links: Lower Layer Protocols for st http://www.darpa.mil/TTO/solicit/BAA07- Autonomous Constellations , 1 Joint 31/F6_BAA_Final_07-16-07.doc Space Internet Workshop, Greenbelt, Maryland, USA, November 13-16, 2000 6. "DARPA awards contracts for fractionated spacecraft program ". 17. Hogie K., et al., Using Standard Internet http://www.darpa.mil/body/news/2008/F6 Protocols and Applications in Space , .pdf. Retrieved 2009-08-15. Computer Networks, 47, 2005, pp.603– 650 7. Mathieu, C. and Weigel, A.L., Assessing the Flexibility Provided by Fractionated 18. Wood L., et al., IPv6 and IPSec on a Satellite Spacecraft , Proc. of AIAA Space 2005 in Space , Proc. of the 58th International Conference, AIAA-2005-6700, Long Astronautical Congress, IAC-07-B2.6.06, Beach, CA, September 2005 Hyderabad, India, September 24-28, 2007 8. Brown, O., et al., Fractionated Space 19. Ivancic W., Internet Technologies for Space Architectures: Tracing the Path to Reality , Applications , DARPA Fractionated Proc. of the 23 rd Annual AIAA/USU Spacecraft Workshop, Colorado Springs, Conference on Small Satellites, SSC09-I-1, August 3-4, 2006 Logan, Utah, US, Aug 10-13, 2009 20. Wood L., Data Routing and Transmission 9. Molette, P., et al., Technical and Economical Protocols , DARPA Fractionated Spacecraft Comparison Between a Modular Workshop, Colorado Springs, August 3-4, Geostationary Space Platform and a 2006 Cluster of Satellites , Acta Astronautica 12 21. Kiefer K., Real-World Experience in Wireless (11), 1984, pp771–784 Instrumentation and Control Systems , 10. Rooney K., An Exercise in Spacecraft Mission CANEUS “Fly-by-Wireless” Workshop, Fractionation , Presentation, Boeing, March 27, 2007 August 2006. Retrieved on 2008-09-01. 22. Hamann R.J., et al., Delfi-C3 Preliminary 11. Mathieu, C. and Weigel, A.L., Assessing the Mission Results , Proc. Of the 23rd annual Flexibility Provided By an On-Orbit AIAA/Utah State University Small Satellite Infrastructure of Fractionated Spacecraft , Conference, SSC09-IV-7, Logan, Utah, Proc. of the 56th International USA, August 10-13, 2009 Astronautical Congress, IAC-05-D3.3.01, 23. "Wireless electricity specification nearing Fukuoka, Japan, October 17-21, 2005 completion ". PCWorld. 2009-08-18. 12. Mathieu C. and Weigel A.L., Assessing the http://www.pcworld.com/article/170360/w Fractionated Spacecraft Concept , Proc. of ireless_electricity_specification_nearing_c AIAA SPACE 2006 Conference & ompletion.html?loomia_ow=t0:s0:a41:g2: Exposition, AIAA-2006-7212, San Jose, r18:c0.016236:b27254916:z0. Retrieved CA, September 19-21, 2006 2009-08-20. 13. Brown, O., et al., System Lifecycle Cost 24. Miura T., et al., Experimental Study of Under Uncertainty as a Design Metric Rectenna Connection for Microwave Encompassing the Value of Architectural Power Transmission , Electronics and Flexibility , Proc. of AIAA Space 2007 Communications in Japan, Part 2, Vol. 84, Conference & Exposition, AIAA-2007- No. 2, 2001 6023, Long Beach, CA, September 18-20, 25. Landis G., Applications for Space Power by 2007 Laser Transmission , SPIE Optics, Electro- 14. Brown, A.W., et al., Software Cost Estimation optics & Laser Conference, Los Angeles for Fractionated Space Systems , Proc. of CA, January 24-28 1994 AIAA SPACE 2008 Conference & 26. Landis G., et al., Space Transfer With Exposition, AIAA-2008-7760, San Diego, Ground-Based Laser/Electric Propulsion , CA, September 9-11, 2008 Proc. of the 28th SAE, ASME, and ASEE

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 10 Joint Propulsion Conference and Exhibit, 38. Golding R., Wong T., Adaptive distributed AIAA-92-3213, Nashville, TN, July 6-8, computing for fractionated space systems , 1992 DARPA Fractionated Spacecraft Workshop, 27. Lafleur J.M. and Saleh J.H., Feasibility Colorado Springs, August 3-4, 2006 Assessment of Microwave Power Beaming 39. LoBosco, D.M., et al., The Pleiades for Small Satellites , Proc. of the 6th Fractionated Space System Architecture International Energy Conversion and the Future of National Security Space , Engineering Conference, AIAA 2008-5714, Proc. of AIAA SPACE 2008 Conference & Cleveland, OH, July 28-30, 2008 Exposition, AIAA-2008-7687, San Diego, 28. Lafleur J.M. and Saleh J.H., System-Level CA, 9-11 September 2008 Feasibility Assessment of Microwave 40. Maessen, D.C., et al., Preliminary Design of Power Beaming for Small Satellites , the Dutch-Chinese FAST Micro-Satellite Journal of Propulsion and Power, vol.25 Mission , Proc. of the 4S Symposium, no.4, 2009, pp976-983 Rhodes, Greece, May 26-30, 2008 29. Jamnejad V. and Silva A., Microwave Power 41. Maessen D.C., et al., Increasing System Beaming Strategies for Fractionated Performance and Flexibility: Distributed Spacecraft Systems , Computing and Routing of Data within the Proc. of the 2008 IEEE Aerospace FAST Formation Flying Mission , Proc. of Conference, Big Sky, MT, March 1-8, the 59 th International Astronautics 2008 Congress, IAC-08-D1.4.6, Glasgow, 30. Turner A.E., Power Transfer for Formation Scotland, September 29 – October 3, Flying Spacecraft , Proc. of AIAA SPACE 2008 2006 Conference & Exposition, AIAA- 42. Guo J., et al., Status of the Fast Mission: 2006-7493, San Jose, CA, September 19- Micro-Satellite Formation Flying For 21, 2006 Technology, Science and Education , Proc. 31. Coulouris G., et al., Distributed Systems: of the 60th International Astronautical Concepts and Design , 4th Edition, Addison Congress, IAC-09-B4.2.5, Daejeon, Korea, Wesley/Pearson Education, June 2005 October 12-16, 2009 32. Tomayko J.E., Computers in Spaceflight: The 43. Bentum M.J., et al., A Novel Astronomical Application for Formation Flying of NASA Experience , NASA Contractor Report th 182505, March 1988 Miniaturized Satellites , Proc. of the 60 International Astronautical Congress, IAC- 33. Meltzer M., Mission to Jupiter: a History of 09-A3.4.3, Daejeon, Korea, October 12- the Galileo Project , NASA SP 2007-4231, 16, 2009 NASA, 2007 44. Muylaert J., et al., QB50: An International 34. Townsend J., et al., Effects of a Distributed Network of 50 CubeSats in LEO for Lower Computing Architecture on the Emerald Thermosphere and Re-Entry Research , Nanosatellite Development Process , Proc. rd th The 23 Annual Conference on Small of the 14 Annual AIAA/USU Conference Satellites, CubeSats Developer’s on Small Satellites, Logan, UT, August, Workshop, Logan, UT, USA, August 8-9th, 2000. 2009 35. Palmintier B., et al., Distributed Computing

on Emerald: A modular approach for Robust Distributed Space Systems , Proc. of the 2000 IEEE Aerospace Conference, Big Sky, MT, March 2000. 36. Palmintier B., et al., A Distributed Computing Architecture for Small Satellite and Multi- Spacecraft Missions , Proc. of the 16 th Annual AIAA/USU Conference on Small Satellites, Logan, UT, August, 2002. 37. Jallad A.H. and Vladimirova T., Distributed Computing for Formation Flying Missions , Proceedings of the 2006 IEEE/ACS International Conference on Computer Systems and Applications, Dubai/Sharjah, UAE, March 8-11, 2006

Paper IAC-09-D1.1.4, 60 th International Astronautical Congress, Daejeon, Republic of Korea, October 12 – 16, 2009 11