Exploiting Social Network Characteristics for Efficient Routing in Ocean Vessel Ad Hoc Networks.

Qihang Bing,Ruobing Jiang,Feng Hong

DOI: https://doi.org/10.1109/ipccc47392.2019.8958733

2019-01-01

Abstract:Communication in the ocean vessel wireless ad hoc networks which currently rely on shore-based cellular stations, is expensive and can only be used near the coast. It is a big challenge for ocean vessels to achieve economic and large-scope communications, due to the rare encounter opportunities and highly dynamic vessel mobilities. Previous works on routing algorithms for vessel wireless ad hoc networks did not take into account the social characteristics of ocean vessel networks. According to our empirical study on real trajectories of more than 7,000 ocean vessel recorded by the Vessel Monitoring System (VMS), we have made innovative observations. Firstly, there exist stable social connections among those vessels frequently encounter with each other. Moreover, stable social communities can be discovered for those ocean vessels with long-term connections. Inspired by the significant observations, two unicast routing approaches based on social society characteristics are proposed for wireless ocean vessel ad hoc networks, i.e., familiarity based routing (FBR) and community based routing (CBR). The key metric for message forwarding decision making of FBR is the expected multihop encounter probability of the potential forwarder with the message receiver. The expected multi-hop encounter probability is derived from the raised definition of familiarity, which measures the connection strengths between vessels. The proposed CBR forwards messages based on the metric of inter-community betweenness centrality of a potential forwarder, which describes the bridge capability between two communities. Extensive trace-driven simulations based on real ocean vessel trajectories are performed and the evaluation results demonstrate the significant performance achieved by the proposed approaches. The transmission cost has been largely reduced with comparable delivery ratio and delay.