Research On P2P Networks Oriented SLBLTrust Model And Its Security Extension

In the past few years, P2P technology has rapidly become the focus of Computer Science industry, with the networks applications from the center server rearranged to the terminal client sides. But the feature of P2P systems, such as open nature and user anonymity, leads to the flood spread of computer virus, useless data, and inauthentic information, because of some irresponsible peers. Therefore there is much attention paid to how to effectively control these dishonest peers' malicious behaviors.An effective solution, proposed by lots of research, is that setting up the trust model in the P2P environment. The existing proposals mainly include two categories: global and local trust model. By analyzing, it is found that the global model introduces complicated computing and heavy communication traffic into the application, while the local way also has the traffic issue.This paper presents a novel local trust model SLBLTrust based on distributed structured P2P topology networks and Subjective Logic quantified computing model. SLBLTrust can efficiently deal with the malicious behaviors mentioned above, and at the same time can precisely estimate whether or not to trust another peer. By reasonable trust storing schema, this model not only can bring the computing complexity and communication traffic overhead under control, but also can be anti-attack. In the end of the paper, by simulating some experiments, it is able to show that the phenomenon brought by the dishonest peers is virtually depressed when using SLBLTrust. And this model's superiority is shown by comparing it with PeerTrust.