The Research On Trust Model Of P2P Network

Peer-to-Peer networks technology has developed rapidly recently, which has been used widely in the fields of file sharing, electronic commerce, and instant messaging and has become hotspot in field of computer network technology research. However, due to its characteristics of autonomy, dynamics and heterogeneity, there exist more and more misbehaviors of malicious peers severely endangering P2P services, which have brought about great challenges to security issue. So, how to establish trust relationship among peers and to manager the P2P networks effectively is an important issue. After in-depth research of trust valuation in P2P networks, this thesis presents a new trust evaluation model based on direct trust value and recommendation trust value.This thesis introduces structure, application, security problems and trust mechanism of P2P networks. On the basis of reading mass references and analyzing the existing trust model, the trust model of the thesis's absorbs the advantages of the trust mechanism in the real-life world, evaluates peer's trust value via quantitative method. It divides peer's trust value into direct trust value and recommendation trust value. The direct trust value is obtained through interactive evaluation of history records between peers. When calculate direct trust value, a dynamic time fading function is introduced to enhance trust value's time-effect. The model negates the hypothesis in that the more trust value the more commendatory credibility, the commendatory credibility is introduced as the weight of commendatory peers, and updated through the evaluation differences between source peers and commendatory peers. A dynamic weight is introduced to calculate comprehensive trust value, so with the number of interaction is increased, the peer is more faith in the direct trust value. Finally, virtual reputation score attribute is added to every peer as incentive mechanism, it can enhance the peer's positivity of participation.The trust model as a module of P2P system, and it doesn't affect the structure of the original P2P implementation. It consists of three parts:trust management, message positioning and transmission and local database. This thesis describes the recommendation mechanism, trust value storage solutions, distributed algorithm and specific flow of work. In the last, we compare our trust model with classic trust model on the basis of the simulation platform. Experiment results show that is accurate to calculate comprehensive trust value and shows advantages on the ability of resisting malicious peers'attack. All those above prove that our trust model has good security and effectiveness.