ICPGF:An Industrial Control Protocol Format-Aware and Feedback-Guided Fuzzing
Xuejun Zong, Bowei Ning, Guogang Wang, Kan He, Lian Lian, Yifei Sun · 2023
To address the issues of low efficiency and poor protocol extensibility in current fuzzing frameworks, this paper proposes a universal fuzzing method for industrial control protocols, ICPGF. Innovatively, we adopt the bootstrap voting expert algorithm to extract the unique format and semantic features of the protocols. By combining optimized mutation strategies and state-feedback mechanism, we guide the generation of testcases and the fuzzing process. The aim is to efficiently discover vulnerabilities in industrial control systems. In a realistic industrial scenario of attack-defense range, we conduct fuzzing on control systems from multiple manufacturers. In comparison experiments with general fuzzing frameworks like AFL and Peach, ICPGF outperforms AFL and Peach in terms of fuzzing efficiency, branch coverage and exception triggers numbers. This fully demonstrates the potential of ICPGF to be an efficient and universal solution in the field of industrial control vulnerability discovery.