Research on Fuzzing Test Case Generation Method Based on Gaussian Mixture Model
Jingqi Fu, Cao Yifan, Yi Mubai · 2019
Using fuzzing test technology to mine vulnerabilities in industrial control networks is one of the important means to ensure the safety of industrial control networks. It is the key for fuzzing test to generate appropriate test case. The traditional fuzzing test framework based on vulnerability library requires a lot of manual participation with low level of automation. In this paper, a fuzzing test case generation method based on Gaussian mixture model is proposed. By learning the characteristics of exception data, the method mixes multiple Gaussian distributions to approximate the actual complex probability distribution function of exception data and uses random sampling to generate new test case. This method can generate data that meets the characteristics of exception data, and the automation level is higher. Through the experimental analysis of Modbus/TCP protocol, it is found that the exception data generation rate of this method is higher than the traditional method.