Research On Trustworthy Mechanism Of Edge Nodes Computing Results Based On Trust Evaluation

The introduction of edge computing technology in industrial networks can solve the problems of equipment request delay,cloud storage and computing overload,and network transmission bandwidth pressure.Security is the key to large-scale applications of edge computing.Among the security issues of edge computing,trusted computing is one of the most important issues.How to ensure the authenticity and trustworthy of the computing results of edge nodes is a challenge.Based on this,this thesis designs a trustworthy judgment method for edge node computing results based on trust evaluation,and builds a test platform to verify the method.The main work is as follows:1.The security requirements of industrial edge computing are summarized.Combined with the security considerations of the edge computing security white thesis,the security requirements and possible attacks of industrial edge computing are analyzed.In order to establish a foundation for building an active defense system for industrial edge computing,an industrial edge computing security architecture with trust evaluation function(TESA)is proposed.2.Aiming at the problem of attackers intercepting and tampering with the transmitted information,based on the industrial edge computing security architecture,a secure transmission mechanism suitable for industrial edge computing is designed.Data source authentication is designed for field devices,and an elliptic curve proxy digital signature process is designed for the communication between the edge side and the industrial cloud to provide complete and reliable data for subsequent trust evaluation.3.Aiming at the problem of maliciously reporting incorrect calculation results after edge nodes are invaded,combined with various security threats faced by industrial edge computing,a trust evaluation-based method for judging the credibility of edge node computing results is designed.The method is divided into the initial trust evaluation stage before the network operation and the trust update stage after the network operation.The trust evaluation mechanism combines the fuzzy evaluation method and the entropy method,and integrates the three-dimensional evidence data of the three effective factors of the direct trust value,the historical trust value and the punishment or reward factor to measure the final trust value of the edge node.And based on the trust value,the malicious edge node detection mechanism is designed.4.Using the industrial wireless system and its edge gateway developed in the laboratory,a test verification platform was built to verify the trust evaluation mechanism proposed in this thesis.The system has been tested and analyzed from two aspects of system function and performance.The test results show that the proposed trust evaluation mechanism is feasible and can evaluate the initial trust value of edge nodes and periodically update the trust value.When the untrustworthy threshold is 0.791 and the credible threshold is 0.991,the detection effect is good,and malicious edge nodes can be detected in time,and legal edge nodes will not be misjudged as malicious edge nodes,which can ensure that the edge side outputs credible computing results.