Modeling and Dynamics Analysis of P2P Networks Based on Evolutionary Games

Abstract

Cooperation has played a key role for good service performance in P2P network. However, without a proper incentive mechanism such cooperation cannot be maintained because nodes avoid spending their resources to share files. In this paper, we employ evolutionary game theory to analyze the dynamic evolutionary process of the mechanism. The game models are constructed based on the stag hunt game (SHG) and snowdrift game (SG). The performance analysis of these two models shows that the system's evolutionary direction depends on the nodes' payoff matrix as well as on the differences of service capacity among nodes. Replicator dynamics and simulation results show that the cost-to benefit ratio is the key factor to increase the proportion of sharing nodes regardless of which game model is used. The results suggest conceptual guidelines to design a good incentive mechanism in the future.