The Research Of Trust Policy Based On The Analysis Of Individual Behavior By Using Game Theory In P2P Networks

P2P network is an open and dynamic computing environment, the autonomy and no central of the node makes it difficult to guarantee the traditional system reliability. Aimed at P2P features establishing trust mechanism is an effective way to protect its security. However, existing studies are focused on trust evaluation and trust management, and lack of research on how to improve trust environment. After analyzing the previous studies on the basis of trust environment, for the shortcomings of the current trust model, established trust model of single transaction and multiple transactions and evolutionary trust model of group transactions, from analyzing a single transaction to multiple transactions, then studied the evolution of the trust between groups of the peers, form individual optimal to overall best. This article the main work includes:1. First, we studied a single transaction between two nodes on the P2P networks, established a single transaction model of the node on the P2P network and identified the optimal trading strategy of a single node. The optimal strategy is trading under a single optimal strategy, but it is not the optimal strategy of a group and how to promote individual and group best is the goal of our study.2. Secondly, we researched multiple transactions between the two nodes on the P2P networks and established a node-based trading model based on repeated games. We try to find a strategy that allows nodes in multiple transactions to maximize its income by the monomer. Strategy selected by the performance of our analysis, we found that model selection strategies with a variety of strategies have shown good performance, with a certain degree of adaptability.3. Finally, we studied the transactions between groups of nodes on the P2P network. Repeated transactions between two nodes model can help us to adjust their strategies to maximize their own income, but individual optimal does not represent the best overall system optimization. Therefore, we consider established a trust model based on groups, by changing some of the factors which make the interests of the individual in the pursuit while maximizing overall system to maximize. In this way, the trust environment will develop in the direction toward the positive. It is same to say, the probability of the nodes on the P2P network select trust increased. So we can achieve the purpose of improving the trust environment.For the research of trust in P2P systems, this article form single to multiple, form individuals to groups, and gradually researched the best benefit strategy, strategy and the node itself form the proceeds of the two groups on individual earnings and overall system nodes earnings were searching process. In practice, these two models can be used in combination to enable them to better address the issue of trust in P2P networks, and promote the sound development of the network.