Adaptive Localization Algorithm For Wireless Sensor Networks Based On Trust Evaluation

Due to rapid development of wireless sensor networks(WSNs),WSNs are widely applied to many fields,such as environmental monitoring,logistics scheduling,and disaster warning.Wireless sensor networks have a random deployment characteristic,which makes the localization process more vulnerable to various attacks.With the aim to solve this problem,this thesis presents the corresponding secure location algorithm and verifies its effectiveness by experiments.The main contributions of this thesis refer to the following:(1)In order to overcome a high system overhead and to prevent a few types of attack,a secure localization algorithm based on trust evaluation is proposed.In this algorithm,the concept of trust management is applied to WSN security localization.In addition,the threshold is analyzed to ensure a high accuracy of attack nodes filtering.Finally,the attack nodes are filtered with a low system overhead to ensure system security.(2)In order to solve the problem of a large overhead in trust management system when nodes move,a secure localization algorithm based on dynamic trust management is proposed.In this algorithm,the trust management is divided into three stages: initial stage,establishment stage,and evolution stage.Simultaneously,the attack behavior is abstracted as an event in order to accelerate the iterative process of trust value through finite state machine.Finally,in the case that anchor node moves,the malicious nodes are identified by iterative trust value,thus,the security of localization process is ensured.(3)In this thesis,a design of adaptive security positioning system,which combines secure localization algorithm based on trust evaluation and the secure localization algorithm based on dynamic trust,is presented.The program code is programmed into nodes,and the nodes are deployed in the area of 10 m x 10 m.The proposed system works in two cases,when node is stationary and when node is mobile.The comparison of positioning error has shown that proposed system can identify malicious nodes and reduce system overhead.Consequently,this thesis proposes a security-positioning scheme for different application scenarios.The simulation results have shown that the proposed algorithm reduces system-positioning error significantly in the presence of attacks.Moreover,the simulation and experiments on system prototype have shown that the proposed algorithm can defend against illegal attacks and that it uses less system resources than AR-MMSE.