Research On Trust Model In P2P Network Security

P2P network has occupied an important place in Internet network, it is concerned by more and more people for its no centralized characteristics. However, the P2 P network itself and its characteristics, such as high dynamic, distributed, autonomous, self-organization, anonymity and so on, provide convenience for poor peers’ malicious and selfish behavior in networks. Because a large number of poor peers exist, the P2 P network suffers greater security risks. The existing P2 P networks lack of effective security mechanism. It is an important subject in the research of P2 P network that how to establish the effective trust relationship among peers. This paper conducts the thorough research in the trust relationship between the P2 P network users, and on the basis of previous studies proposes a P2 P trust model—CRRPTrust(Classification Recommend and Reward Punish Trust), which is based on classification recommend, reward and punishment mechanism.CRRPTrust model have drawn on the trust relationship in human society, and used the trust relationship in the peers of the model. Peers earn evaluation from other peers through providing resources. We can calculate the trust degree through these transaction records. This paper will adopt the quantitative method to assess the peers trust degree.Firstly, this model groups peers according to the similarity degree of peer’s resources type, peers in the same group have similar resources. In this model, the paper proposes the calculation rules of the group trust degree, trust degree between two groups and user’s trust degree between groups recommended. The trust degree between peers can be calculated through the above three trust degrees using the formula giving in this paper, besides, the central peer can calculate the global trust degree of the certain peer according to the trust degree between peers.Secondly, this paper introduces the incentive and punishment mechanism, the central peer assigns the corresponding virtual currencies for each peer in the P2 P network according to the global trust degree. When the peer obtains resources successfully, it needs to pay the virtual currency. When the sharing resources of the peer are successfully downloaded, the peer’s virtual currencies are increased. The change of virtual currencies is the standard to solve peers global trust degree.Finally, the model is implemented by programming. According to simulation, this paper contrasts the model before and after the grouping and introducing the incentive mechanism, and compares this model to Engin Trust in some respects. Experiments show that the model reduces the number of messages and the searching time of the P2 P network in certain extent, and is able to effectively curb the malicious peer’s bad behaviors.