The Research Of P2P Trust Model

P2P network is dynamic, self-organized, and anonymous and has some other features. P2P network can not guarantee that all the peers can provide reliable resources and good service. Some peers even have malicious behavior, such as provision of false information, spread of illegal advertising, dissemination of trojans and worms. In order to solve these problems, the trust mechanism is introduced to P2P network and trust model is built to establish trust relationship between peers.After the study of existing P2P network trust models, analysis of fundamental issues related to trust research and the discussion of the related key techniques of P2P trust model, a relatively new P2P trust model——P2P-BMRTM (Bayesian and Multiple factors based on Reputation Trust Model) is presented.BMRTM comprehensively considers all aspects of P2P application and classifies the parameters according to Bayesian Network design idea. The parameters are designed based on Multi-factors Trust Model while the parameters are expanded and the computation formula is improved. The similarity of P2P network and human society is considered. The Direct Trust and the Recommended Trust is distinguished.Recommended Trust is weighted by the recommendation credibility to calculate comprehensive recommended trust. Finally, the final trust of target peer is integrated by direct trust and comprehensive recommended trust.A variety of simulation experiments are designed and executed. The results show that P2P-BMRTM can effectively improve the overall P2P application network performance in transaction successful rate, and can effectively reduce P2P network hazard by the malicious peer's exaggerated behavior and slander actions.