Design and Analysis of a Bluetooth Hotspot & Wi-Fi Hotspot

ISSN 2319-8885 Vol.04,Issue.15, June-2015,

Pages:2871-2876

www.ijsetr.com

Design and Analysis of a Bluetooth Hotspot & Wi-Fi Hotspot 1 2 SHASHIKANT B. AHIRE , TUSHAR S. PAWAR 1Assistant Professor, Dept of MCA, A.C.Patil College of Engineering, Kharghar, Navi Mumbai, MH, India, E-mail: [email protected]. 2PG Scholar, Dept of MCA, A.C.Patil College of Engineering, Kharghar, Navi Mumbai, MH, AP, India, E-mail: [email protected].

Abstract: Now a days we can use internet by using Wi-Fi in the mobile phones too. But, mobile phones with Wi-Fi are very expensive. But Bluetooth being common and efficient way of transferring of data, can as well be used to gain internet access. Using Bluetooth Hotspot we can use internet, on mobile phones with Bluetooth, from other phones which will host their own internet connection. This will make available the provision of accessing the internet for a large percentage of mobile phone users even if they do not have GPRS connection or Wi-Fi facility. Wireless hotspot networks have become increasingly popular in recent years as a means of providing Internet access in public areas such as restaurants and airports. In this paper we present the first study of such a hotspot network. We examine five weeks of SNMP traces from the Verizon Wi-Fi HotSpot network in Manhattan. We find that far more cards associated to the network than logged into it. Most clients used the network infrequently and visited few APs. AP utilization was uneven and the network displayed some unusual patterns in traffic load. Some characteristics were similar to those previously observed in studies of campus WLANs.

Keywords: Hotspot, Bluetooth, Internet, GPRS or Wi-Fi Facility.

I. INTRODUCTION internet made available to him by the host phone. The host A. Hotspot will set up an internet connection and share the service with Wi-Fi hotspots are very common now days. Laptops and the public. Interested client can request the host to receive PDAs use these hotspots to access the internet as these access to the service. The host will accept or reject the devices have Wi-Fi facility built into it. But mobile phones connection. If accepted the client can surf the internet using do not have Wi-Fi facility in them. Currently many mobile the host‟s internet connection free of cost. companies are providing models with Wi-Fi with smart phones, but still those models are costly. But lower end C. Wi-Fi Hotspot models also come with Bluetooth now. So it is a better idea In recent years, deployment of Wireless Local Area to create a Bluetooth hotspot using which mobiles with Networks (WLANs) has boomed as demand for wireless Bluetooth can access the internet. With this technology there Internet access grows and IEEE 802.11 technology matures. is no need for the mobile phones to have a GPRS connection 802.11 WLANs can now be found in offices, homes and or even a SIM card in it to access the internet. Bluetooth campuses. One increasingly-popular use for 802.11 provides a way to connect and exchange information between networking equipment is to provide wireless „hotspots‟, that devices. Bluetooth is a radio standard and communications is, providing wireless Internet access in popular public places protocol primarily designed for low power consumption, with such as airports, shops and caf´es. An understanding of how a short range (power class dependent: 1 meter, 10 meters, these hotspot networks are used can guide network design, 100 meters) based around low-cost transceiver microchips in hotspot deployments, and the development of technologies to each device. be used on WLANs. In this paper we present one of the first studies of a deployed 802.11 hotspot network. We collected a B. Bluetooth Hotspot network activity trace lasting approximately five weeks from Bluetooth hotspot application allows a smart phone to host the Verizon Wi-Fi HotSpot network. We analyze the network a Bluetooth hotspot which can be accessed by other phones in terms of users, Access Points (APs) and traffic, and with Bluetooth. By this we are making the access of internet compare some of our findings with those for a college on mobile phone available cheaply to majority of mobile campus wireless network and a corporate wireless network. phone users who do not have a Wi-Fi service in their phones. In the next section, we review related work. In Section 3, we The hosting smart phone can get the internet connection from describe the study environment and in Section 4 we describe a Wi-Fi server or via GPRS of their own service provider. the tracing methodology. Section 5presents the most The client phone need not spend any money to access the interesting features of the data and compares them to results

SHASHIKANT B. AHIRE, TUSHAR S. PAWAR obtained in previous studies of WLAN usage. In Section 6 setting. Kotz and Essien examined a college campus wireless we formulate our conclusions. network when it was first installed in 2001. Henderson et al. returned to the same network after it had matured in II. LITERATURE SURVEY 2003/2004. Two other campus WLANs that have been A. History of Hotspot studied include the University of North Carolina and the A hotspot is a site that offers Internet access over a University of Saskatchewan, while Balazinska and Castro wireless local area network through the use of a router analyzed usage of a corporate WLAN . While hotspots are a connected to a link to an Internet service provider. Hotspots popular topic in both the business and research worlds, we typically use Wi-Fi technology. Hotspots may be found in are unaware of any other papers that examined a deployed coffee shops and various other public establishments in many hotspot network. Balachandran et al. examined the challenges developed urban areas throughout the world. Public access facing hotspot networks, while Verhoosel et al. proposed a wireless local area networks (LANs) were first proposed by generic hotspot business model. Henrik Sjödin at the Net World+Interop conference in The Moscone Center in San Francisco in August 1993. Sjödin did III. EXISTING SYSTEM not use the term hotspot but referred to publicly accessible A. Wi-Fi wireless LANs. Most widely used wireless technology today, is Wi-Fi. Wi- Fi hotspots are deployed in shopping malls, cafes, public B. Bluetooth libraries etc. It gives internet access to Wi-Fi enabled laptops Bluetooth is a short range wireless communication standard or to Wi-Fi enabled mobile phones for free or at a low cost to operating in the 2.4 GHz hand, designed to connect mobile the ones in that hotspot area. A wireless network uses radio phones, laptops, PDAs and other portable devices [3]. It is a waves, just like cell phones, televisions and radios do. In fact, cable replacement standard for a Personal Area Network communication across a wireless network is a lot like two- (PAN). Its compact, light and affordable features allow way radio communication. people to travel conveniently, reducing the problems of cable connections which take up luggage space and get tangled B. The Study Environment of Wi-Fi quickly. Bluetooth deploys a Frequency Hopping Spread Network: The Verizon Wi-Fi HotSpot network (VWHN) Spectrum (FHSS) technique to improve its immunity from consists of 312 APs distributed around the island of interference. This ensures that data packets are not lost and Manhattan. 1 APs are installed in the ceilings of consequently, increases the performance of the network. Verizonowned phone booths. Each AP is a Proxim OriNOCO AP-2500 802.11b AP2, enclosed within a weatherproof box C. Bluetooth Technology Overview containing the AP, a DSL modem, a power regulator, and an The basic components of any Bluetooth application consist external antenna. APs are connected to the Internet by a 1.5 of the following items: Mbps downstream and 768Kbps upstream ADSL connection.  Stack initialization In the weatherproof boxes, the APs have a maximum range  Device management of close to 300 feet but in practice, due to environmental  Device discovery interference, an AP‟s effective range is approximately 150  Service discovery feet. Although all APs share the same SSID, the VWHN does  Service registration not support roaming between APs. When moving from one  Communication AP to another within the network, a user must reauthenticate to obtain Internet access at the new AP. D. Challenges for Bluetooth Hotspot The execution of such a heavy application on the mobile Users: The VWHN is currently provided solely as an phone effectively and efficient facilitation of data transfer amenity service to Verizon Online (VONL) DSL and dial- up between two mobile phones. Also we have to enable the customers. Customers of these services use their VONL application to access the internet connection on one mobile username and password to log on to the network. As of phone from any possible source so we have to include December 2004, 10,511 unique. options for choosing from a list of possible sources for a mobile phone and all that sources should be able to share Authentication, Authorization, and Accounting: To obtain their internet connection through the Bluetooth with the Internet access at a Verizon Wi-Fi HotSpot, a user must first client‟s mobile device. Also the communication between the log into Hotwire, a proprietary hotspot management system client and the server over the Bluetooth is a challenge. Both developed within Verizon. To log in, a user first associates to the mobile phones will have to be synchronized so that the the AP and opens a web browser, which is redirected to a client can send data in the form of page requests and it can in web page requesting a username and password. Access is turn receive data in the form of browser readable pages from granted upon submitting a valid username and password. the internet through the serve. Prior to login, an associated user‟s Authentication, Authorization, and Accounting (AAA) state is considered E. Wi-Fi Hotspot pending at the AP. After login, it is considered valid. A user Recent studies have characterized wireless network usage in may also have an unknown AAA state before sending any a variety of environments. Tang and Baker studied a packet packets to the AP. A user in this state is treated as a pending radio network composed of nearly 25,000 radios distributed user because they have similar access privileges. across three major metropolitan areas. Balachandran et al. analyzed WLAN usage over three days in a conference International Journal of Scientific Engineering and Technology Research Volume.04, IssueNo.15, June-2015, Pages: 2871-2876 Design and Analysis of a Bluetooth Hotspot & Wi-Fi Hotspot C. Bluetooth Hotspot Service Using A PC and A Dongle the entire trace is implied. Valid cards have unrestricted There are two major components of the entire system. access to the Internet at the AP where they are valid. The first component is the Bluetooth hotspot network which consists of many machines in a client server configuration. Pending Card: A card in a pending AAA state during a The Bluetooth network is essentially a collection of many given time period at a given AP. Pending cards have Internet Bluetooth devices like mobile phones and PDAs. A„*server” access limited to certain VWHN-related websites. A valid connects the entire network to the internet acting as a card is not guaranteed to be seen as pending even though it gateway to the “clients”. The second part of the network is must have been pending at some point prior to login. Note the Bluetooth service manager. This is mainly the software that the set of pending cards is not disjoint from the set of used to manage the users of that particular system. This valid cards. system takes care about the login and authentication. The customer enters his/her user details to gain access to the Session: A session begins with the appearance of a card at an service. Then the authentication process validates the users. AP in a given AAA state (valid or pending), and ends when User details are stored in a database with access details like a the card is either no longer at the AP or when the card username and password. A centralized database will maintain changes AAA state. the details of all users registered to the system. Once the user enters the hotspot, a simple “login” facility will enable the Active AP: An active AP is an AP to which one or more user to access the Internet. cards are associated (regardless of the cards‟ AAA state) during a given time period. IV. FEATURES OF HOTSPOT There are two sides of Bluetooth Hotspot application, Valid AP: An AP at which one or more associated cards was server side and the client side. The server side program run seen with a valid AAA state during a given time period. on a mobile phone does the work of hosting the internet Pending Traffic: Traffic generated by pending sessions. connection via Bluetooth. This side uses the existing internet connection of the phone to host the internet service. The Valid Traffic: Traffic generated by valid sessions. program shows requests from client and gives the user a choice to allow or block a particular client. Once the client is Inbound: Traffic sent by the AP to the card. selected, the program listens to the localsocket where the Bluetooth connection with the client is established for any Outbound: Traffic sent by the card to the AP. requests from the client. When the server gets the page request from the client, it forwards the request to the World V. SCOPE BLUETOOTH HOTSPOT Wide Web server to get the appropriate response. This As the number of Bluetooth products increases each year, response in the form of web page is then forwarded to the it is important to develop applications and services to take client over the Bluetooth connection established. The client full advantage of their potential and capabilities. A side program first searches for the nearby Bluetooth devices broadband hotspot is one application where Bluetooth has a hosting the service and populates the list with the names of value in providing Internet access to mobile users. the Bluetooth devices. The connection is made to the server Consumers owning a Bluetooth enabled mobile phone can by clicking on the connect button. Then the browser can be easily access a Bluetooth hotspot to browse the Internet used to access web. Any request that is made by the browser without having to carry a PDA or a laptop. Today, the is listened to by the client side program at localsocket. When majority of people are in possession of a mobile phone. The the client side program is run a Bluetooth connection is number of mobile phone users is increasing each year. established with the hotspot. So once a request is obtained Mobile phone creates more possibilities for social from the browser the client program forwards the request to networking. The deployment of Bluetooth hotspots will the hotspot through the Bluetooth connection channel and widen access to broadband services using mobile phones not waits for the response. . Once the response is obtained from only to professionals and mobile workers but more the hotspot the client returns the data to the browser through importantly, to other segments of consumers who want to go the connection made on the localsocket. on-line for non-business related and sociable purposes, and to consumers who do not own PCs. Users will also be able to IV. DEFINATION OF Wi-Fi share their internet connection through their own phone. And AAA State: The Authentication, Authorization, and people who do not have an internet connection will not need Accounting state of a card at a given AP. A card may have a to use a computer or activate GPRS on their own phone but valid, pending, or unknown state. A card has an unknown can use the internet connection of their friend who has state before sending any packets to the AP. Hereafter, we use Bluetooth on their phone and also has Bluetooth hotspot. the term pending to describe both the pending and unknown states because cards with these states have similar access VI. DESIGN privileges and we treated them as the same. The architecture of Bluetooth Hotspot consists of the following modules and shown in Fig.1: Card: A wireless NIC, identified by MAC address. 1. Server Mobile: Valid Card: A card in a valid AAA state during a given time  Internet Browser period at a given AP. If no period is specified, the period of  Wi-Fi card/ GPRS card  Bluetooth International Journal of Scientific Engineering and Technology Research Volume.04, IssueNo.15, June-2015, Pages: 2871-2876

SHASHIKANT B. AHIRE, TUSHAR S. PAWAR 2. Client Mobile: browser also displays the page on the device when the  Internet Browser Bluetooth responds with the requested page.  Bluetooth Bluetooth: The client pairs its Bluetooth device with the 3. Internet server‟s mobile using Bluetooth. The page request made by the browser is received by the Bluetooth and it sends this request to the server‟s Bluetooth over established Bluetooth connection between client and the server. The Bluetooth also receives the page sent by the server‟s Bluetooth and forwards the page to the browser of this device.

C. Internet The internet is the World Wide Web which receives all the page requests from the server via Wi-Fi/GPRS card. It analyzes the request and responds to the server with appropriate response in the form of a web page.

VII. RESULT OF WI-FI HOTSPOT Over the 36-day trace period (which includes 22 complete days of data), we gathered 746,397 SNMP records. We saw 26,925 total cards, of which 1,682 were valid at one point in Fig.1. Architecture of Bluetooth Hotspot. the trace. We summarize our results in a manner that facilitates comparing the VWHN with WLANs studied in A. Server Mobile other environments. In addition, we investigate usage The server mobile is the mobile which hosts the internet characteristics of the VWHN that differ from previously service for the client to use it. studied networks.

Internet Browser: The internet browser on the server mobile A. Users is used by the server himself to access the internet. The user For a WLAN such as Verizon‟s, understanding the user is types a domain name into the address bar in the browser and critical to building and maintaining a successful network. the page request for that page is sent by the browser to the Card Activity: Patterns in the number of valid cards for each Wi-Fi/GPRS card. The browser also displays the page day of the study strongly mirror the number of pending cards received from the World Wide Web server, through Wi- on the network for each day of the study (Fig.2). Some users Fi/GPRS card, as a response to his request. have multiple sessions in a day, and so we observe approximately twice as many sessions as cards. Wi-Fi/GPRS Card: The server mobile accesses internet with the help of Wi-Fi or GPRS. It uses Wi-Fi card or GPRS card for the same. These cards forward the requests made by the browser to the World Wide Web server via the link established between the mobile and the internet. It also accepts the response received from the World Wide Web server and recognizes it as response to the server‟s request or client‟s request and accordingly forwards the page to the browser or to the Bluetooth respectively.

Bluetooth: The Bluetooth on the server side does the work of accepting page requests from the connected client‟s Bluetooth and forwards this request to the Wi-Fi/GPRS card. Also it sends the response received from the card back to the client‟s Bluetooth. This interconnection of Bluetooth requires pairing of the Bluetooth devices to establish the trust of the client with the server.

B. Client Mobile The client mobile is the mobile which uses the service of Bluetooth hotspot provided by the server. The client accesses Fig.2. Cards and sessions per day. The cards and sessions the internet with the help of Bluetooth hotspot. for a day appear just to the right of its tic mark. Blank spaces represent holes in the data. Sundays are labeled. Internet Browser: The client after pairing his device with The x-axis is on a log scale. server‟s mobile, uses the internet browser to access the A puzzling question is raised by the small number of valid internet. The client enters the URL into the address bar in the cards (1,682) in comparison to total cards (26,925) seen browser. The browser makes a page request for the URL and during the trace. Why did so many cards associate to Verizon sends the request to the Bluetooth of his own device. The International Journal of Scientific Engineering and Technology Research Volume.04, IssueNo.15, June-2015, Pages: 2871-2876 Design and Analysis of a Bluetooth Hotspot & Wi-Fi Hotspot APs but not log in (and attain a valid AAA state)? Perhaps advantage of the network will do so. Further investigation of some users are simply curious and select the VWHN SSID the data shows that the greatest number of pending cards was when they see it is an available network, or perhaps some seen at APs in the Midtown area, a mostly business district, clients‟ wireless networking management utilities chose to while valid cards were heaviest at APs in the Upper West automatically associate to the network. Side, a residential area.

B. Access Points Busiest periods: The hotspot APs were not particularly busy, We had 282 APs respond to SNMP polls. We now look in even during peak usage periods. The greatest number of detail at the AP statistics. simultaneous valid sessions ever hosted by an AP was 7, whereas the most cards ever simultaneously associated to an Activity: Examining AP activity over the course of the trace, AP was 24. The most valid cards seen by an AP during a day some APs see many cards while others see relatively few was 10, and the most pending cards ever seen by an AP (figure not shown). In testing for linear correlation, the during a day was 106. On the Dartmouth campus, in contrast, proportion of variation in valid cards that is explained by the the maximum simultaneous users on one AP is 89, and the linear regression of valid cards on pending cards (r2) is only maximum cards seen on an AP in a single day is 405. 0.391. In other words, a device‟s association with Verizon‟s WLAN poorly correlates with the likelihood Log-log CCDF Traffic: Most APs see little traffic, but several see significant (Complementary Cumulative Distribution Function) of valid amounts. This pattern is similar to the traffic pattern across session durations APs on a college campus with APs handling traffic more unevenly than on a corporate WLAN. Pending traffic was mostly inbound (83.23%) although there are high outbound loads on some days (Figure 11). Pending traffic accounted foronly 2.07% of total traffic. But this small percentage still totaled a median of 0.29 GB each day, which could become expensive for a hotspot provider who is paying for upstream bandwidth that is being consumed by non-customers (i.e., pending cards). Hotwire access logs show that HTTP requests from automated processes (e.g., Windows Update) being redirected to the Hotwire login page generated much of the pending traffic.

VIII. CONCLUSION With the use of internet so rampant in these days the access of internet must be simplified and made common. Bluetooth Hotspot is the perfect way to realize this need. Bluetooth Hotspot already exists for use in computers and in Fig.3. Scatter plot of pending cards at an AP and valid laptops. But to carry around a laptop everywhere to access cards at an AP. free internet on mobile phone is a lot of effort. Also a Bluetooth dongle should be used. That is why Bluetooth Hotspot will be of a great help. It is easily installed on to the mobile phone and does not require any extra hardware for it to run. This will help a lot of people who do not use internet so frequently but require it on the go. The modules and design for this project have been conceptualized and the hardware and software requirements for development and usage have been analyzed and documented successfully. This paper presents the first analysis of a production 802.11 hotspot network. We examine five weeks ofSNMP traces rom the Verizon Wi-Fi HotSpot network in Manhattan. We find hat most users access the network infrequently, but daily, weekly, and hourly trends still emerge. Far more cards associate to the network than log in, and it is difficult to explain why.

The vast majority of cards spend most of their time at a single AP, and few cards even visit more than one AP. APs Fig.4. vary widely in their utilization. Most APs were active on any given day, but fewer saw a login. The number of cards that of that device actually using the WLAN. Perhaps this reflects associated to an AP is a poor predictor of the number of users an uneven distribution of VONL customers around the city as that logged in. Most network traffic was caused by valid shown in Fig.3. Or it might be that an AP‟s surroundings play sessions and in particular by fewer than 5% of valid users. a role in determining whether or not someone able to take Traffic varied across days and exhibited unusual hourly International Journal of Scientific Engineering and Technology Research Volume.04, IssueNo.15, June-2015, Pages: 2871-2876

SHASHIKANT B. AHIRE, TUSHAR S. PAWAR characteristics. We intend to look further into similarities between the hotspot network data and previously-collected campus datasets. Hotspot data is somewhat harder to obtain than campus WLAN data, and our conclusions in this study we were limited by the absence of data concerning what users were actually doing on the network along with the coarse granularity of SNMP polls. It would be useful to understand what aspects of a hotspot network can be simulated or modeled using campus WLAN data.

XI. ACKNOWLEDGEMENT The authors are grateful to Conor Hunt, Sean Byrnes, Paul Perry and the other members of Paul Perry‟s team at Verizon who allowed this study to take place. The authors also thank Mike Leahy of Verizon Data Services for his help in ollecting the data.

X. REFERENCES [1]http://developers.sun.com/mobility/midp/articles/bluetooth 1/. [2]“Widening Access to Broadband Hotspots Employing Bluetooth” by Jazilah Jamaluddin, Ratish Nair, Reuben Edwards, and Paul Coulton. [3]“Spontaneous Hotspots: Sharing Context-Dependent Resources Using Bluetooth” by Xavier Le Bourdon, Paul Couderc, Michel Banâtre [4]“Secure Cognitive Mobile Hotspot” by Zaheer Ahmed, Habibullah Jamal, Rizwana Mehboob, Shoab Khan, and Muhammad Shahbaz [5]“Anatomy of a Bluetooth Hotspot” by Chris De Herrer . [6]T. Henderson, D. Kotz, and I. Abyzov. The changing usage of a mature campus-wide wireless network. In Proceedings of Mobi-Com 2004, pages 187–201, Philadelphia, PA, Sept. 2004. [7]D. Kotz and K. Essien. Analysis of a campus-wide wireless network. Wireless Networks, 11:115–133, 2005. [8]M.Papadopouli, H. Shen, and M. Spanakis. Characterizing the duration and association patterns of wireless access in a campus. In 11th European Wireless Conference, Apr. 2005. [9] D. Schwab and R. Bunt. Characterising the use of a campus wireless network. In Proceedings of INFOCOM 2004, pages 862–870, Hong Kong, China, Mar. 2004. [10] D. Tang and M. Baker. Analysis of a metropolitan-area wireless network. Wireless Networks, 8(2–3):107–120, Mar.- May 2002. [11]J. Verhoosel, R. Stap, and A. Salden. A generic business model for WLAN hotspots: a roaming business case in The Netherlands. In Proceedings of WMASH 2003.

International Journal of Scientific Engineering and Technology Research Volume.04, IssueNo.15, June-2015, Pages: 2871-2876