Research On Optimization And Generation Method Of Test Case For Industrial Communication Protocol Based On Fuzzing Test

At this stage,Industrial Control Systems(ICSs),as the management and control system in the country's important infrastructure,play an indispensable role in the industrial production and manufacturing process.When carrying out various industrial control activities,industrial communication protocols are often regarded as an important medium for realizing real-time control and remote management between different industrial equipment.However,with the integration of information technology and industrial manufacturing,the inherent security flaws of industrial communication protocols are gradually exposed,which makes industrial control systems face more and more network attack and intrusion methods,and traditional IT security testing techniques are not well suited for industrial control systems.Therefore,based on the security defense requirements of industrial control systems,this paper takes the industrial communication protocol as the research object,and proposes a test case optimization method based on the Fuzzing test framework to find more vulnerabilities with a certain number of test cases,but only finding the optimal test cases cannot uncover all potential vulnerabilities as much as possible.In order to achieve higher test case coverage,this paper builds an improved GAN test case generation engine to generate a large number of new test cases with strong similarity to the optimized test cases,which achieves the effect of improving the coverage of test cases,and finally achieve high coverage and high efficiency in mining vulnerabilities for industrial communication protocols.First of all,based on the Fuzzing test framework,this paper proposes an optimization method of test cases for industrial communication protocol based on the improved genetic algorithm.Furthermore,this method designs a new individual selection strategy as a selection operator,which can actively participate in the test case optimization process.Specifically,in this individual selection strategy,a selection operation based on populations with high and low fitness is introduced to enhance the diversity of individual selection,which can not only search for a batch of test cases with higher fitness values in the iteration of the algorithm,but also further improves the efficiency of vulnerability mining in test cases.Secondly,based on the above optimized test cases of industrial communication protocols,this paper proposes a method for generating test cases of Generative Adversarial Networks(GAN)based on parameter optimization.Moreover,this method uses GAN as a test case generation model,and optimizes the parameters of GAN through an improved Particle Swarm Optimization(PSO)algorithm.Finally,this paper realizes a PSO-GAN test case generation engine based on multi-parameter dynamic optimization,which can achieve better test case generation effect and generate a large number of new test cases that are more similar to the optimized test cases,thereby enhancing the coverage of test cases.Finally,the simulation experiment environment of industrial communication protocol is built,which can further evaluate the effect of the Fuzzing test case optimization method and the test case generation method designed in this paper.The experiment results can be concluded that,on the one hand,compared with other test case optimization methods,the proposed test case optimization method makes the test case have higher test efficiency;on the other hand,compared with different parameter optimization methods of test case generation engines,the proposed test case generation method not only has a higher generation effect when generating test cases,but also has more efficient training efficiency in the parameter training process.