The Research Of Dynamic Trust Evaluation In Mobile Ad Hoc Networks

Mobile Ad Hoc networks is made by a team of nodes which with wireless communication transceiver. It is no infrastructure.Each terminal network node can move freely,has equal status and don’t need the support of the existing basic network facilities.It can be constructed fastly at any time and any place.Because of using radio channel and distributed control technology,and node power supply is limited.It is more vulnerable to passive eavesdropping,initiative invasion,denail of service,sleep deprivation and other network attack.The packet transmission between nodes faced serious challenge.Many traditional security strategy and mechanism of the traditional network is not applicable to the network.As the premise and foundation of network security,trust,through the measurement and evaluation of trust relationship between the nodes,establish a relatively flexible safety measure mechanism. At present,the trust evaluation has become a hotspot in network security.Because of the uncertainty of the trust itselt and unique feature of mobile Ad Hoc networks,constructing the trust evaluation model which with high precision and low burden is still a challenging task.In the mobile Ad Hoc networks,we studied the dynamic trust evaluation in the plane structure and hierarchical structure.The main work of this paper is as follows:At first,in the plane structure, dynamic trust evaluation is proposed based on the credibility method. In the existing model of trust evaluation,there are some defects on evidence collection and integration.So a trust evaluation model of multidimensional evidence fusion is proposed.First,detecting the various historicl behavior of nodes througn the expanding Watchdog to wide the dimension of direct evidence collecting.Then,each behavior is calculated by using bayesian mehtod which meet the beta distribution. Through the credibility method,all behaviors are merged together to improve the accuracy of assessment of the trust.Finally,the simulation results demonstrate the effectiveness of this model.Analysis shows that recommended weight is quantitavely evaluated by information similarity measure,reducing the impact of malicious recommendation,ensuring the trust assessment robustness.Secondly.in the hierarchical structure,a trust evaluation is proposed based on the transferable belief model.At present,there is a less of trust evaluation model in the hierarchical structure.At the same time,they didn’t consider the role of nodes in the trust evaluation of hierarchical structure.This model,according to different nodes role,using the transferable belief model,assigned different belief function and established the different trust relationship vector.First,the malicious recommendation was defensed by using the confidence.Second,the different belief functions of trust relationship vector were merged by using Dempster rules and realized the transfer of trust between clusters.Belief function distribution and belief updating is difficult,we solved the problems perfectly.