The Research On Trust-based Intrusion Detection In Wireless Sensor Networks

The rapid development of wireless communication technology,sensor technology and embedded computing technology gave birth to a new type of network--wireless sensor network(WSNs).As a kind of wireless network,wireless sensor networks have broad application prospects in military,environment monitoring and industrial production and other fields.However,due to the openness of the wireless communications spectrum and harsh environments in which nodes deployed makes wireless sensor networks exist big security risk,which restricts the range of applications of wireless sensor networks.Existing security strategies are basically to ensure network security from the perspective of prevention,but it cannot prevent malicious attacks which come from internal.As a reasonable complement of security and defense mechanism,the main purpose of the intrusion detection technology is to find any malicious nodes which attempt to undermine the confidentiality and reliability of network.Through the detection and isolation of malicious nodes existing in the network,a secure communication environment can be built and the normal operation of the network can be ensured.We first studied the existing intrusion detection technology for wireless sensor network carefully,analyzed their implementation principle and the existing shortcomings.Then,by simulating trading evaluation mechanism in human society,we proposed a new wireless sensor network intrusion detection algorithm based on trust.Take the existence of intelligent attack into consideration,some trust status of evaluation node often appear uncertain through the judgment of the trust model.To solve this problem,we proposed a mechanism to accelerate trust status judgment,which will find the malicious node existing in network as soon as possible.The main works of this paper include the following aspects:First,combined with the characteristics of wireless sensor networks,we proposed an intrusion detection method based on trust.Through monitor the communication behavior of node from multiple aspects,a trust model is built for the evaluated node.And then combining the fuzzy theory and evidence theory to deal with the collected behavior characteristics and obtained the trust value of the evaluated node.Finally,according to the trust value of the evaluated node,the trust status of the node can be determined.Experimental results show that the proposed trust-based intrusion detection algorithm has a high detection rate and reliability compared to the previous algorithm,which can identify the presence of malicious nodes in the network and prevent the destruction of malicious nodes on the network.Secondly,in view of the uncertain state of evaluated node,we proposed a mechanism to accelerate trust state judgment.Through analyzing the history assessment record of evaluated node and the distances between collected real-time behavior evidence,each of behavior evidence was assigned a weight value,and then modifies the evidence according to the weight value and reuses the trust model to make an evaluation on the evaluated node again.This method is simple and effective.By using the interval time between two rounds of trust evaluation to make an assessment on evaluated node will not increase the communication burden on the network,and can effectively shorten the survival time of the malicious node in the network.The experimental results show that,after adding the mechanism to accelerate trust state judgment,the detection rate of proposed intrusion detection system rises faster and has stronger intrusion tolerance ability than previous scheme.