A Novel Coverage-Guided Greybox Fuzzing Method based on Grammar-Aware with Particle Swarm Optimization

Abstract

Coverage-guided Greybox Fuzzing (CGF) as a popular testing approach has been widely used in software testing. However, the existing CGF has some problems, for example, the testing efficiency is often poor in the face of structured input. To solve this problem, Grammar-Aware Greybox Fuzzing (GAGF) has gained attention for its use of abstract syntax trees (AST) to help processing the structured inputs and it has achieved higher fuzzing efficiency than CGF. However, the improvement of efficiency may not be enough. Therefore, we proposed a particle swarm optimization algorithm to help GAGF to further improving the efficiency. The proposed algorithm can selectively optimize the mutation operator in GAGF mutation stage, as well as accelerate the mutation efficiency of fuzzing to achieve more higher code coverage.