Design And Implementation Of Intelligent Detection Scheme For Protocol Vulnerability In IIoT

With the application of industrial Internet more and more widely,its security problems have gradually begun to pay attention to,before the attacker to find its potential security risks is very important.Fuzzy testing is a common,simple and efficient tool for vulnerability detection.It is also widely used in industrial Internet vulnerability mining at present.In the traditional protocol fuzziness testing method,generating test cases requires a lot of manpower and time to analyze the protocol specification,and with the change of the protocol,generating methods will not be universal.Therefore,it is necessary to design a fuzzy testing method which can reduce the cost of protocol analysis and obtain better testing results.In recent years,with the excellent performance of machine learning methods in various fields,fuzzy testing field also began to use it for intelligent testing.However,at present,machine learning methods are seldom used in the field of industrial control fuzzy testing,and the test results of generated test cases are not ideal.Therefore,based on the text generation antagonism network,this paper proposes a test case generation method for industrial control protocol that can solve the above problems,in which two kinds of text generation antagonism network models are adopted,and it is more in line with the requirements of fuzzy testing in the preservation of generated data.In addition,a fuzzy test system for industrial control protocol based on the test case generation method mentioned above is implemented.In order to prove the effectiveness of the proposed method,an experiment is carried out on Modbus TCP protocol,a common industrial control protocol.The experimental results show that compared with the original GAN test case generation model and the traditional fuzzy test tool Peach,the proposed model is better in terms of the test case pass rate,vulnerability mining efficiency and test case diversity.