Research on Intelligent Fuzz Testing Technology for Power Internet of Things Terminal

For many power terminal devices, firmware binary code can be extracted directly from the device by different methods. And the binary firmware program security test, so as to find the Internet of Things terminal device security vulnerabilities. In this paper, the intelligent fuzz testing technology for power Internet of Things terminal is studied. Firstly, the basic framework of binary dynamic piling is studied, and the bottleneck of the existing framework is analyzed. In order to improve the performance of firmware program in dynamic pile driving environment, a dynamic simulation and dynamic pile driving framework combining user level simulation and system level simulation are proposed. Based on this framework, the pile driving method oriented to program coverage and the pile driving method oriented to program vulnerability guidance are studied respectively, which provides support for cross-platform fuzz testing. Secondly, dynamic stain analysis techniques are studied from both offline and online aspects. The advantage of offline dynamic smudge analysis is that it can fine-grained how data is propagated in memory during program execution, and can assist in the generation of initial test seeds. The advantage of online dynamic blot analysis is that it does not affect the performance of simulation test, and can dynamically infer the relationship between input bytes and constraints, so as to guide fuzz testing to carry out targeted variation. Finally, the intelligent prediction model of program vulnerability points is studied to intelligently locate potential defects in binary firmware programs. In order to solve these potential defects, the fuzz testing based on program coverage and potential vulnerability guidance is used, and the particle swarm optimization algorithm is used to optimize.