Mobile social networks : system and applications

Abstract

Mobile social networks (MSNs) enable the interactions between people through their mobile devices within a social networking environment. More and more MSN applications, services and systems are being proposed and developed, changing the way people communicate with each other and enabling ubiquitous collaboration among people. In this thesis, we make three contributions to the development of future MSNs. First, we present a social network framework based on an innovative concept called social vectors. Basically, a social vector can be used to define and quantify the social relationship between two social entities. In general, social vectors can also be used to facilitate the development of MSN system and applications. Second, based on social vectors and inspired by previous work, we present an opportunistic routing protocol called Hybrid Opportunistic Routing for Social Entities (HORSE). HORSE seeks to combine the advantages of the previous opportunistic routing protocols in order to forward social messages effectively and efficiently over MSNs. In particular, it is inspired by the "six degrees of separation" theory, which provides the basis for forwarding social messages through a chain of friends. Simulation results using real-world data traces are presented to evaluate the performance of HORSE as compared with other schemes. The results indicate that HORSE can achieve high delivery ratios while maintaining low overhead costs. Third, based on social vectors, we present two related applications. The first one is a Smart Shopping System, which identifies customers using RFID technology and finds similar customers using social vectors with the purpose of providing customized marketing messages for customers. The second one is a Dynamic Signage System with a social vector-based dynamic playlist mechanism, which enables the display of dynamic contents that matches with people interests. Simulation results are presented to show the effectiveness of this mechanism.