An Overlapping Community Detection Algorithm for Opportunistic Networks

A more detailed community structure can contribute to a better understanding of the network, which can also benefit efficient routing protocols and QoS schemes designing. For an Opportunistic Network which consists of different kinds of mobile nodes, its topology changes over time. Therefore the community detection becomes more difficult than static situations. Moreover the overlapping community detection is a more complex problem. This paper analyzes the time varying topology of Opportunistic Networks and the overlapping community structures of human. Then, we propose a new detection algorithm to solve the overlapping community detection problems in Opportunistic Networks. Only with the local network topology information and a short period, nodes can get their overlapping community structures by our detection algorithm. Numerical simulations with both scenarios of movement models and real trace data are presented to illustrate the accuracy and efficiency of our algorithm.