Efficient Routing Algorithms Combining History and Social Predictors in Mobile Social Networks

Abstract

A mobile Social Network (MSN) is a type of wireless networks formed by people moving around carrying mobile devices. In this paper, we specifically study the MSNs that are formed impromptu, e.g. when people gather together for a conference, event, or festival. We refer to them as Impromptu Mobile Social Networks (IMSNs), which allow people to communicate in a lightweight fashion based on contact opportunities via local wireless bandwidth. In IMSNs, node connections are time-dependent and short-term. The existing MSN routing algorithms using network analysis of social network graphs and static node social features may not be suitable for IMSNs. Thus, we propose novel hybrid routing algorithms based on two time-related factors node contact history and dynamic social features to capture node mobility in IMSNs. We first propose a hybrid algorithm called Hisso that makes a weighted combination of the history and the social predictors based on these two factors. And then we upgrade it to Enhanced Hisso by introducing a novel concept called social circle in the social predictor to improve message delivery. Simulation results comparing our algorithms with the existing ones and with the ones that only consider one factor show that our algorithms outperform the others in terms of delivery ratio and latency with a slight increase in the number of forwardings. The results also confirm the effectiveness of using a node's social circle in our algorithms.