H-Fuzzing: A New Heuristic Method for Fuzzing Data Generation

摘要

How to efficiently reduce the fuzzing data scale while assuring high fuzzing veracity and vulnerability coverage is a pivotal issue in program fuzz test. This paper proposes a new heuristic method for fuzzing data generation named with H-Fuzzing. H-Fuzzing achieves a high program execution path coverage by retrieving the static information and dynamic property from the program. Our experiments evaluate H-Fuzzing, Java Path Finder (JPF) and random fuzzing method. The evaluation results demonstrate that H-Fuzzing can use fewer iterations and testing time to reach more test path coverage compared with the other two methods.