Research On Key Technologies Of P2P Trust Model

In recent years there has been an increasing focus on P2P network, P2P network has been applied in many fields, and there is still big room for the development for P2P network. However, the inherent characteristics of P2P network, such as openness and anonymity, provides an opportunity for the malicious acts which have a serious affect on the development of P2P network. Currently, the construction of trust relations among of nodes becomes a research focus to deal with the security problems in P2P network.First of all, a further study was performed on trust mechanism, related concepts of trust model and some key technologies on the construction of trust model. Furthermore, STRW algorithm was presented to solve the network cost problem of flooding algorithm which is widely used on the basis of key research on trust value search algorithm of current trust model. The algorithm uses K-Walker Random Walk algorithm which is based on Random Walk theory to search trust value which greatly reduces the network cost in the process of trust value search. At the same time the reduction of searching range may decrease the accuracy of the search results. In order to solve this problem, the searching process is performed on the double dynamic trust network topology which is made up of direct trust relationship and recommender trust relationship. The K Walk Random Walk algorithm is improved by using Preferred Walk instead of Random Walk which greatly raises the accuracy of searching result. Moreover, the quick link consisting of recommender relationship reduces the search path and increases the average search speed.Finally, it is proved by the algorithm performance analysis of STRW that its network cost is far less than flooding algorithm. In addition, a simulation experiment to verify the accuracy of searching result of STRW is performed, the result of which indicates that the accuracy of search result is not decreased by the reduction of the searching range, furthermore, the ability of resisting malicious false recommendation is enhanced by STRW algorithm.