Research Of Trust For Nodes In P2P Networks Based On Repeated Game Theory

P2P technology has been widely applied to the Internet with the rapid development of networks, and P2P application has attracted a lot of users. It allows peers in P2P network connect others'computers directly. While peers interact all need through suppressive servers in traditional C/S network. With the scale of the network, the burden of the server will be heavier, once the server poly may cause the network paralysis. P2P networks have no suppressive server; each node is both the client and the server.The distributed feature of P2P networks has brought a series of new problems and challenges. Because of lack of centre governing and the characteristic dynamic equality between each node of P2P application, the existing threat of network basically are: bandwidth consumption issues; Contents identify issues, the issues of trust, Can management issues and so on. The trust problem is the most prominent problem; we can analysis node in P2P network with the behaviors of game theory.In order to research P2P trust game model, this paper firstly introduces the basic concept of P2P networks and P2P network model, and analyzes the safety problems exists in P2P network. Then introduces the basic concept and basic elements of game theory, and do a detailed description about the Nash equilibrium and repeated games. Aiming at the problem of trust P2P networks, this paper combines with the game theory to study the behavior strategies of the nodes in P2P network, and puts forward four kinds of P2P trust game model: a repeated game based on the P2P node behavior strategy; unknown node types of P2P trust game model research; based on supervision game of incentive mechanism of P2P nodes; the game of P2P between the nodes Nash equilibrium. Use the related theoretical model and mathematical formula of game theory, establishing P2P trust models in various situations based on game theory, and analyses the behaviors of nodes to obtain optimal incentive strategies.This article validates the feasibility of the model with simulation tools named Gambit, and the model really can effectively promote P2P networks to be safe and stable, and can cause network nodes consciously to join in P2P networks, reducing the node take malicious attacks.