Research On Trust Model Of Wireless Sensor Network Based On Recommendation And Risk Control

Wireless Sensor Network(WSN)is a key technology of the Internet of things,which is widely used in many fields.WSN is characterized by a large number of sensor nodes and large network scale,and is usually deployed in exposed external environments,so it is vulnerable to various forms of attacks.Authentication is the basis of network security.Traditional authentication mostly adopts centralized authentication,but this authentication mechanism has many disadvantages: the security of the network is completely dependent on the authentication center,once the authentication center is maliciously attacked,the whole authentication system will collapse;When the network scale is expanded,the performance of the network will be affected due to the limited computing capacity and storage capacity of the certification center.In addition,with the development of WSN,the network topology is ever-changing,and the centralized authentication is not flexible enough.Therefore,centralized authentication has low security and poor scalability.This security problem can be effectively solved by establishing a distributed trust model in WSN.This paper studied the distributed trust model in detail,and proposed the trust model of wireless sensor network based on recommendation and risk control(RRCTM model).The specific work and innovation contents are as follows:Firstly,in order to calculate the trust value of the interaction node more accurately,this paper introduces the recommendation mechanism.It not only obtains the direct trust value according to the historical interaction record between nodes,but also entrusts the third-party node to recommend the target node,integrates the trust data of itself and the third-party node,and comprehensively evaluates the trust value of the target node.In this paper,the recommendation mechanism is designed in detail.Secondly,in order to ensure the accuracy of the data transmitted in the network,a risk control mechanism is set up in the model.This paper refers to the Va R risk measurement method in the financial field and proposes a risk control model.The risk is measured according to the interaction in the network,and then the trust value is further adjustedaccording to the risk value,so as to improve the sensitivity of the system and effectively resist the attack of malicious nodes.Thirdly,in order to ensure the real and safe nodes,a reward and punishment mechanism is set up in the model.The third-party node should be responsible for its own recommendation behavior.If the trust data provided is true,it should be rewarded with corresponding credit.Otherwise,the corresponding credit penalty is applied to make the model have strong dynamic adaptability.Simulation results show that the RRCTM model has low cost and accurate trust evaluation,which can effectively resist the attacks of various malicious nodes in the network and guarantee the security of wireless sensor network.