Research On New Anonymous Trust Model Based On Resource Evaluation For Peer-to-Peer Networks

With the constant development of information technology, peer-to-peer as a kind of advantageous and significant network structure has become an indispensable network application system. In the field of file sharing, distributed area and other related fields, P2P has been widely applied and has a larger development space, especially in the field of network file sharing, openness, autonomy and anonymity which makes data share more efficient. However, the higher openness led to the falsity of network resources, the abuse of network data and severe malicious attacks, eventually led to the low credibility of resources.In order to solve the problem such as resource abuse and security issues in P2P, I introduced the concept of trust model to prompt P2P system homemade formula to establish effective evaluation mechanism, by measuring the credibility of the resource sharing nodes, thus ensuring the high reliability of the network.Through thoroughly researching the existing file sharing P2P trust models, carefully analyzing the variety of trust mechanism of trust models, and fully considering the advantages and disadvantages of typical trust models, then the paper proposes a high reliability and high efficiency P2P trust model.Compared with the previous interaction evaluation based on sharing node and the general trust model which only considers sharing node as the computing subject, the model proposed in this paper transfers the computing subject from sharing node to share the resource itself. Through calculating the satisfaction value of the resource itself and reversing effect to the Shared file nodes, the model effectively solves the limitations of resources’ quality unilaterally decided by sharing nodes. In terms of computational efficiency, through comprehensively considering the effective evaluation feedback of historical evaluation behavior of nodes and the comprehensive satisfaction value of shared resources, it quickly gets nodes’ reliable reputation, and reduces the complexity of iterative arithmetic on traditional trust model.In the aspect of safety, because of the transfer of computing subject, part of the concept of reliability is transferred to resource itself. At the same time, by improving and implementing the mechanism of CAN(Content Addressable Networks), combining the ideology of MIX which economizes the relation of model trading, it also limits the spread of false and malicious resources fundamentally, effectively against common malicious attacks in resource sharing environment.The model also puts forward the effective incentive mechanism, to encourage network nodes to actively involve in resource sharing, at the same time to give effective punishment to the malicious behavior, to well restrain the spreading of false resources in the network. By means of the simulation experiment, they fully proved the performance improvements in efficiency and safety aspects of the model in this paper.