The Research Of Trust Evaluation Model In Ad Hoc Network

Due to self-organization, no center, the limited resources of nodes, topology changes dynamically and completely open radio channel characteristics; Ad-Hoc networks are vulnerable to malicious nodes’ attacks, such as uncooperative, malicious recommendations, and other forms of collusion attacks. Ad-Hoc network security issues directly affect the network robustness and lifetime of the network. Trust model is an important method to solve the current Ad-Hoc network security issues, and how to be able to accurately describe the trust relationship between the nodes is the primary problem that any trust management model must face. In this paper, we study trust of Ad-Hoc networks in two aspects: behavioral surveillance and trust assessment.In Ad-Hoc networks nodes are usually leading to its strategy to become selfish or malicious node because of their limited resources. If nodes just enjoy the services provided by other nodes and there exists a lot of selfish nodes in the self-organizing network long-term, the nodes will provide reliable service without enthusiasm, mobile ad-hoc network will ultimately unsustainable. The malicious nodes which provide unreliable service selectively with strategic policy will also affect the quality of service Ad-Hoc network. Therefore, designing an effective mechanism to make nodes provide reliable services in MANET and improve the operational capacity of the network is the key to solving the problem.Due to the lack of centralized server, the node cannot be calculated by a trusted third party in Ad-Hoc network. Because changes in the network topology or wireless conflicting, evidence sample space for trust evaluation may be incomplete or unreliable, affecting the effectiveness of trust assessment. This paper mainly discusses the trust evaluation model of Ad-Hoc networks, the main results are as follows:(1) During the updating process of direct trust, use the trust vector similarity to determine the weighting factor. Punish malicious nodes and reward loyalty nodes by the use of incentive mechanism. The model makes rewarding efforts greater than punishment. The measures can effectively inhibit the malicious nodes quickly enhance their trust, forcing the malicious node can follow the original intention of the mobile ad hoc networks, provide better services to other nodes. Through the accurately describe of direct trust relationship between nodes proposed the synthesis rules the of direct trust based on relevance and incentive mechanism which will improve the success rate of transaction in the Ad-Hoc network and enhance network security and robustness.(2) Monitoring recommended nodes’ behavior, detecting malicious nodes, building trust rules for recommended trust synthesis node can improve network security and robustness in the face of an insecure network environment. In the trust recommendation process, malicious nodes may launch attacks or defamation collusion to provide false trust information, evidence of conflict between false trust and real trust lead to the synthesis of trust contrary to intuition and unreasonable results. In this paper, we propose a trust evaluation model based on filtering recommendation mechanism and D-S theory’s synthetic rules with a discount factor to monitor the nodes’ behavior and to solve the problem.