Off-campus UMass Amherst users: To download campus access theses, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Theses that have an embargo placed on them will not be available to anyone until the embargo expires.

Access Type

Open Access

Document Type


Degree Program

Electrical & Computer Engineering

Degree Type

Master of Science in Electrical and Computer Engineering (M.S.E.C.E.)

Year Degree Awarded


Month Degree Awarded



Social network, mobile P2P, social trust, wireless ad hoc network, overlay network


Wireless mobile internet market is still an unprecedented, uncaptured territory for cellular service providers. The shortage and high cost of downlink data bandwidth in a cellular network has remained a huge factor for the slow growth of data services in mobile devices. Although there has been a significant evolution in telephony infrastructures in form of 3G and 4G systems, the potential of high speed ad hoc network for sharing cellular spectrum have not been realized to its full potential. Like (e.g. Verizon) users can share voice minutes with friends, there is a potential for sharing the unutilized cellular bandwidth among friends to increase net data speed. In a scenario like a football stadium where people visit in groups, although a lone phone cannot stream a high quality replay video, unused cellular bandwidth of proximate friend’s devices can automatically be used in real time to view the replays. An available secondary ad hoc network such as Wi-Fi or Bluetooth in phone can be used for sharing this cellular bandwidth. Thus, we propose BuddyShare, a novel social-based automatic bandwidth sharing overlay platform on short range ad hoc devices to increase net data speed. The motivation stems from the fact that the location of mobile users tends to be clustered to form “people hotspots” such as conferences, stadiums, stations, buses and trains. For example, in a scenario like a football stadium where people visit in groups, although a lone phone cannot stream a high quality replay video, unused cellular bandwidth of proximate friends’ devices can automatically be used in real time to view the replays.

Our work creates an overlay on horizontal ad hoc network to enable users to form a group among socially trusted members who can collaboratively share their data connections. Social trust is automatically derived from social relationships obtained by mining mobile-phone behavior pattern. This work aims to improve the overall utilization of the data connection, and increase the data rate of individual users without compromising their privacy and unauthenticated usage. The user privacy is preserved by using the bandwidth resources of only socially trusted member of the user, which also guarantees against unauthenticated exploitation of expensive bandwidth. Our proposed work promises to deliver win-win situation to users, content providers and service providers. The advantages of users are: 1) Increased data rate for the same cost.2) Secure and trusted overlay based communication for sharing resources. The advantages for the service providers are manifold: 1) Customer increase: More customers will avail the data plan due to social influence. 2) Customer retention: [18] Customers part of the social-cum-adhoc network are least likely to leave the network.3) Group subscription: Service provider can get bulk subscriptions as collaborative groups increase data rate.

In this work, we address some key technical issues of developing a socially aware overlay collaborating medium. Some of the addressed functionalities associated with the overlay formations are group discovery, creation, management and actual data distribution. This proposal also accounts the computation of social trustworthiness by using standard social networking analytics. We also account the several key technical challenges associated with management of overlay on mobile nodes and trust computation using abstract social network. In order to verify the usefulness of BuddyShare, we collected realistic datasets from various sources (questionnaires, mobile device logs, social networking portal) and conducted analyses and simulations on it. The analyses concluded that sample users from the dataset shared sufficient social trustworthiness. The real events from the datasets were captured in the simulations. These simulations showed that, by using Bluetooth as a horizontal ad hoc medium, an user can scale his data speed three times on average for sufficient duration per day.

This thesis achieves the following objectives: 1) It presents a comprehensive design for an overlaid social based internet sharing platform called BuddyShare. 2) It presents a social analysis to validate the concept of social trust among users. 3) It delivers a flexible simulation platform to realistically simulate mobile phones with dual interfaces. 4) It presents the results of simulations of real events captured from the device logs of sample users. These results conclude the usefulness of BuddyShare work.


First Advisor

Aura Ganz