ABSTRACT
Department of Electrical and Computer Engineering, University of British Columbia
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328 11.2 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
11.2.1 Opportunistic Sharing in DTNs/MSNs . . . . . . . . . . . . . . . . . . . . . 331 11.2.2 Mobile Traffic Offloading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 11.2.3 Information/Content Spreading in SNSs . . . . . . . . . . . . . . . . . . . . 334
11.3 Details of TOSS Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 11.3.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 11.3.2 Mobility Impact in the Offline MSN . . . . . . . . . . . . . . . . . . . . . . . 337 11.3.3 Spreading Impact in the Online SNS . . . . . . . . . . . . . . . . . . . . . . . 339 11.3.4 Access Delays of Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
11.4 System Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 11.5 Trace-Driven Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
11.5.1 Measurement of Mobility Impact, IM . . . . . . . . . . . . . . . . . . . . . . . 343
11.5.2 Measurement of Spreading Impact, IS . . . . . . . . . . . . . . . . . . . . . . 343 11.5.3 Measurement of Access Delay, Ai(t) . . . . . . . . . . . . . . . . . . . . . . . 344
11.6 Performance Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 11.6.1 Mapping Schemes of Online SNSs and Offline MSNs . . . . . . 345 11.6.2 Initial Pushing Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347 11.6.3 Satisfying 100, 90, and 80% of Users . . . . . . . . . . . . . . . . . . . . . . 348 11.6.4 On-Demand Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
11.7 Discussion of Practical Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 11.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Rapidly increasing mobile traffic has become a serious concern of mobile network
operators. In order to alleviate this traffic explosion problem, there have been some
efforts to research offloading traffic from cellular links to local short-range communi-
cations amongmobile users in proximity of each other. In this chapter, we first survey
related studies on opportunistic user-to-user (device-to-device) content sharing, and
then present a new proposal to carry out social-aware mobile traffic offloading as-
sisted by social network services via opportunistic sharing in mobile networks, called
TOSS. TOSS initially selects a subset of mobile users as seeds, depending on their
content spreading impact in online social network services (SNSs) and their mobility
patterns in mobile social networks (MSNs). Then users share the content with each
other via opportunistic local connectivity (e.g., Bluetooth, Wi-Fi Direct) according to
their social relationships. Due to the distinct access patterns of individual SNS users,
TOSS further exploits the user-dependent access delay between the content gener-
ation time and each user’s access time for opportunistic content sharing. We model
and analyze the process of traffic offloading and content spreading by taking into
account various options in linking SNS and MSN data sets. We present trace-driven
evaluations to show that TOSS can reduce 63.8 to 86.5% of the cellular traffic while
satisfying the access delay requirements of all users. Thus, SNS-based traffic offload-
ing by opportunistic sharing can provide an effective and efficient content delivery
service that is promising for the future wireless and mobile networks.
