Efficient opportunistic routing with social context awareness for distributed mobile social networks

Mobile social networks (MSNs) are developed from mobile ad hoc networks. Nodes in such networks usually have social characteristics. In recent years, researchers are trying to use the social characteristics of the network to propose new data forwarding metrics, so as to design more efficient routing algorithms. However, most of the proposed algorithms only consider local context information, which leads to the performance of the routing is not optimized enough. In this paper, we introduce two key metrics, namely, social relationship and social activity. The metrics will be used to search the best data forwarding nodes to improve the probability of data delivery. We propose a prediction-based social-aware opportunistic routing (PSOR). In the proposed method, node's social profiles are used to search relay candidates set, and the discrete-time semi-Markov prediction model is used to find the probability distribution of node transition between communities. Many simulation experiments based on real traces show that the proposed PSOR algorithm is more efficient to maximize the packet delivery probability than other state-of-the-art algorithms.