ABSTRACT
CONTENTS 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 9.2 Methodology: A Data-Driven Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
9.2.1 Experimental Data Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 9.2.1.1 CoNext07 [22] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 9.2.1.2 Infocom06 [5] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 9.2.1.3 CoNext08 [29] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 9.2.1.4 Dartmouth01 [16] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 9.2.1.5 Hope08 [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
9.2.2 Evaluation Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
9.3 PeopleRank: A Social Opportunistic Forwarding Algorithm . . . . . . . . . 246 9.3.1 The PeopleRank Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 9.3.2 PeopleRank Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
9.4 Ensuring Fairness in Mobile Opportunistic Networking . . . . . . . . . . . . . 250 9.4.1 The Efficiency Fairness Trade-off . . . . . . . . . . . . . . . . . . . . . . . . . . 250
9.4.1.1 Absolute Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251 9.4.1.2 Absolute Fairness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
9.4.2 Real-Time Distributed Approach for Fairness-Based Forwarding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252 9.4.2.1 Desired Fairness and Satisfaction Index . . . . . . . . . 252 9.4.2.2 The FOG Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 9.4.2.3 FOG Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
9.5 Forwarding in Large-Scale Mobile Opportunistic Networks . . . . . . . . . 257 9.5.1 Forwarding Drawbacks in Large-Scale Opportunistic
Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 9.5.2 Forwarding within Sub-Communities . . . . . . . . . . . . . . . . . . . . . . 258
9.5.2.1 Classification and Forwarding in Sub-Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
9.5.2.2 Impact of Different Community Classifications on Forwarding Performance . . . . . . . . . . . . . . . . . . . . 260
9.5.3 Forwarding across Sub-Communities: The Community-Aware Framework (CAF) . . . . . . . . . . . . . . . . . . . . . 262 9.5.3.1 The Impact of Community Classification on
CAF-Enabled Rank-Based Forwarding Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
9.5.3.2 CAF vs. BubbleRap . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 9.5.3.3 The Impact of MultiHomed Nodes . . . . . . . . . . . . . . 266 9.5.3.4 The Cost of CAF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
9.6 Social-Based Trust in Mobile Opportunistic Networks . . . . . . . . . . . . . . . 270 9.6.1 Social-Based Trust Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272 9.6.2 Relay-to-Relay Trust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272 9.6.3 Source-to-Relay Trust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
9.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
9.1 Introduction The growth of social interaction has evolved from the confined requirement of physical proximity, to telegraph and telephone networks, ultimately exploding to various forms of cyber-based interactions enabled by the ubiquity of the Internet. These social interactions, no longer requiring physical presence amongst participants, are enabled by many applications including email, chat, and more recently, online social
applications create a virtual space through which users build social networks of their acquaintances, where they can freely interact, independently of where they are located. However, when people with similar interests or common acquaintances are within physical proximity of one another, beyond human sensory ranges, they have no automated way to identify potential relationships. The relationship between virtual social interactions and physical meetings remains largely unexplored. Exploiting high-granularity physical proximity and social information has been recently tackled in the domain of mobile opportunistic networks [22, 32].
