Research On Trust Model And Excitation Mechanism In P2P Network

With the development of the Internet, Peer-to-Peer Network (P2P) has become one of the popular research fields nowadays. Making full use of the user resources on network edge to accelerate the search, location and downloading of resources and provide mass data storage, P2P technology has a wide range of applications in the fields of file sharing, distributed computing, cooperative system, electronic commerce, depth search engine, etc.P2P network has broad development prospects, but it faces grand challenges as well, esp. the safety problem. Peer-to-Peer network is anonymous, dynamic and exoteric, which on the one hand offers more convenience to users, but on the other hand poses a potential safety hazard of being destroyed and attacked by hostile users. The main reason for this is that in the early stage of designing Peer-to-Peer network, the factor of safety was not taken into account. The establishment of interactive node trust relation is one part of the safety problem; therefore, it is quite necessary to build up a trust management system in P2P network.Trust reflects the overall evaluation of the behavior and capacity of a peer by another peer. Interactive node trust relation in P2P network is a powerful guarantee for cooperation and resource sharing, which directly affects the overall behavior of the P2P system. The trust problem in P2P network can be solved by establishing unfailing trust management model in the system. The present thesis mainly focuses on the study of P2P network trust model. In reference to the social network trust model, it puts forward a subject-based self-recommendation trust model to deal with the possible threats in the interactive process.The present thesis mainly consists of the following parts:First, it gives a detailed introduction to the basic knowledge of P2P network, analyzes the trust model in distributed environment, and probes into several typical trust models and the research methods related to them.Second, in view of the fraud of some spiteful nodes in P2P network, it puts forward a subject-based self-recommendation trust model. What interests the users more is the node resources, not the node itself; therefore, we can categorize the resources according to subjects and bring forward the thought of subject trust. In terms of computing the node's trust value, we introduce the local trust table and the subject feedback table, improve the computing method of global credibility, and replace the complicated iterated operation with trust value matching to save operation expenses. In response to the attacks of spiteful nodes, the present model analyzes the security from the aspects of cheating behavior and slanderous attacks. We achieve subject clustering and trust model excitation mechanism through the interest option in the subject feedback table. We conducted simulated experiment and performance analysis of the trust model. By establishing the trust mechanism among the nodes in the network, we identify, restrict and stop the behavior of spiteful nodes, weaken the harm that spiteful nodes do to the network, and enhance the security of the network.Third, we provide the mechanisms discouraging free riding in P2P networks. Our solution primarily focuses on locating free riders and taking actions against them. We propose a framework in which each peer monitors its neighbors, decides if they are free riders, and takes appropriate actions.Finally, we simulate and analyze some experiments about the proposals against free riding.