Research On Trust Model In Trusted Network

With the computer technology and internet application developed rapidly, the wayof people’s work, live and play is more and more closely related to the networks,security and trust are becoming increasingly important parts in information system.Meanwhile, due to the inherent fragility of the traditional network systems, unreliable ofnetwork security measures and uncontrollable of user behavior, network security isfacing a huge challenge. Therefore, to build a safe, reliable, credible, controllable andtrusted network is particularly necessary. Trust model as the core of the trusted networkscan fundamentally solve the issue of credibility of the network through the credibilityevaluation mechanisms regulate the behavior of entities, is an effective way to securetrusted networks. Under such background, this thesis does some exploratory research ontrust model and its application.First, fundamental issues related to the evaluation of trust model are discussed and anovel framework of trust model based on identity and behavioral trust is proposed.Second, a new trust model based on identity and behavior assessment termed asdynamic subjective trust model is proposed. It uses direct assessment regard as main,indirect assessment considered as supplement and some additional factors to evaluateentity’s trust. In addition, in order to expand performance of the model, a novel trustprediction algorithm based capacity segment is presented.Third, a trust-based routing algorithm as a practical application of our trust model isproposed. It uses trust relationship as bond to constrain nodes’ packet forwardingbehavior, incentives to regionalize security detection and punitive measures to isolatethe malicious nodes. Our algorithm takes those methods to balance safety and networkperformance and improve the overall performance of routing.Finally, with the simulation environment of OPNET and VS2005, this thesis doessome experiments to analyze the performance of our proposed trust model.