TTL sensitive social-aware routing in mobile opportunistic networks

Social characteristics such as community and centrality have been used as forwarding metrics in literature to predict nodes encounter probability in mobile opportunistic networks. However, the community detection and centrality estimation based on cumulative contact does not take the lifetime of messages into consideration. By investigating the periodicity of social characteristics evolution, this paper presents a novel method to calculate the community and centrality metrics of nodes, in time scales of τ-window and time-to-live (TTL), respectively. Such metrics are then exploited in a proposed social-aware routing algorithm, PerEvo, to predict the encounter probability of nodes during the lifetime of the forwarded message. Extensive trace-driven simulation results show that PerEvo improves the message delivery ratio significantly, while keeping a comparable delivery overhead with the existing social-based routing schemes.