An Evolutionary Game-Based Mechanism for Routing P2P Network Flow among Selfish Peers

Aimed to improving the efficiency of self-interested P2P node’s routing traffic through a congested network and overcoming confusion condition caused by selfish routing in P2P networks, we introduce an evolution game-based routing model to study the selfish routing behaviors of nodes in P2P networks. In the paper, we model the routing behaviors of nodes as a noncooperative routing game, in which self-interested player’s route traffic through a congestion-sensitive network. We extend the model of selfish routing to study the dynamical behaviors of nodes, by adopting a generalized approach of imitative dynamics. We not only analyze the model’s stability and convergence, but reveal that the efficiency of P2P node’s routing can be improved when the model reach an equilibrium state. Finally we also give an algorithm and experiments on how to improve P2P traffic efficiency based on our evolutionary game model.