Based Interest Groups And Recommended Trusted P2p Network Security Model

With more and more attention and research invested into it, the application of Peer-to-Peer network has access into various fields. For its wide use, the space for its development is considerable in the future. Due to its opening up and anonymity, there spring up more and more misbehaviors of malicious peers endangering the regular performance of the network. Thus, how to identify the malicious peers and inhibit their misbehaviors has become the focus of our research in the trust mechanism.After deep research into the development and trust mechanism of P2P network, the recommendation-based trust model is improved with reference to the trust relationship of communication network of human society. The interest group and recommendation-based trust model that is improved transacts with the valuable peers picked out from the entire network by "interest group", shortening the peers transactions and lessening the calculation of trust value and network traffic. This model distinguishes the transaction and recommendation trust value definitely by "recommendation credibility". The recommendation trust value is calculated through the deviation of feedbacks, which avoids iterative computation, simplifies the implementation and resists the attacks of malicious peers effectively. This model calculates trust value of the peers more really and efficiently by balancing the proportion of local trust value to recommendation trust value of other peers in the calculation of trust value. In the end, this paper presents the storage and achievement scheme of the model and the solution tosecurity problems in current P2P network.Through comparison between the improved model and other recommendation-based trust model, the simulation results proved that the improved model has better performance in evaluation accuracy, security, resistance to attacks on the malicious peers and lower operating expenses.