Contacts-aware opportunistic forwarding in mobile social networks: A community perspective

Abstract

Exploiting community structure for opportunistic forwarding decisions in mobile social networks offers a promising paradigm to improve the transmission performance and reduce the extra network overhead. Actually, people will have closer relationships and more opportunities to contact with each other if they are in the same community. In this paper, the activeness of nodes and the probability of reaching the destination are investigated based on the node contacts in the trace. Then the Contacts-Aware Opportunistic Forwarding (CAOF) scheme, which includes inter-community and intra-community phase, is proposed. In the inter-community phase, the node with higher global activeness and source-to-destination probability is selected to serve as the relay. Besides, in the intra-community phase, the forwarding decisions are determined by the local metrics. Furthermore, we compare the proposed CAOF scheme with several benchmark forwarding algorithms, including BUBBLE Rap, SPRINT, Epidemic and JDER. The validity of the modeling and the soundness of the analysis are verified through extensive experiments with real traces, which illustrates that it outperforms other routing strategies in heavy traffic scenarios.