Research Of P2P Trust Model Based On Binary Trust

Peer-to-Peer network, commonly abbreviated to P2P, has developed from the traditional C/S architecture of network topology to without central infrastructure. Its open and anonymous characteristics, make arbitrary nodes are free to join and withdraw from the P2P network, to provide or obtain resources and services. Also for this reason, it makes it more difficult and expensive to maintain or manage the P2P systems. Another reason is that there are many selfish or fraudulent nodes in P2P networks. The main reason of these issues because P2P lacks effective constraint mechanism and trust between nodes, which is the main factors that affect the further development of P2P.One of the most important questions of trust model research is how to search and calculate the trust value of nodes, in other words, how to evaluate the trust values. In the traditional network environment or electronic commerce system, most trust models focus on computing trust value for the providers of resources, but actually trust is subjectivity and complexity. Even if precision of the trust value is relatively high, trust is still uncertain. Therefore, how to improve the reference value of trust for e-commerce transactions, how to resist the attack from malicious nodes, and how to protect the interests of good nodes and improve the enthusiasm of good nodes, should become the focus of P2P Trust Model.To solve the above problems, this paper proposes:the binary representation of the trust, the compensatory mechanism after the failure transaction, and the marking methods of malicious nodes. The model and calculating method of binary trust is introduced, and expressed as (transaction trust, compensation trust). According to small world principle and the binary trust value, the nodes are divided into clusters with the way of one-dimensional priority. Also based on the chord routing method, the search success rate of good node in the transaction is improved. This paper also introduces the way of the compensation transaction, the influence of compensation transaction on transaction trust, and the method to identify malicious nodes. Compensation mechanism restricts the attack of malicious nodes, the influence of compensation transaction on transaction trust improves the enthusiasm of good node to participate in the compensation mechanism. Experimental results show that the proposed model can identify malicious nodes effectively and protect the good nodes. The success rate of transactions between nodes in P2P systems has greatly enhanced.