.png){kind=icon}
.png){kind=icon}
Return
QPSOFuzz: A Fuzzer Integrating Quantum-behaved Particle Swarm Optimization Algorithm and Logistic Mapping
.png){kind=icon}
PRE
AI
DOI：10.3837/tiis.2025.12.019.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Mutation-based grey-box fuzzing has become a widely adopted technique to test software vulnerability. Its effectiveness largely depends on the mutation operator selection strategy. Many fuzzers employed the uniform probability distribution to schedule mutation operators, which was inefficient in practice. In this situation, many schemes adopting adaptive mutation strategies that can dynamically adjust probabilities of mutation operators had been proposed. However, the path exploration ability of some schemes could be further improved. And the resource consumption of some schemes was too high. Thus, in this paper, we propose QPSOFuzz, which is an improvement work of MOPT. QPSOFuzz integrates the Quantum-behaved Particle Swarm Optimization (QPSO) algorithm and Logistic mapping to initialize the parameters, adjust the contraction-expansion factor, and update the probabilities of mutation operators. Besides, QPSOFuzz redefines the global optimal probability, combining both the current efficiency and historical efficiency. We evaluated QPSOFuzz against 3 stateof-the-art fuzzers across 9 real-world programs. The extensive evaluation results show that QPSOFuzz could achieve higher path coverage while keeping lower resource consumption. And in certain specific scenarios, only QPSOFuzz could still trigger crashes, while the other three fuzzers failed to trigger any crashes.
Keywords:
Grey-box Fuzzing
Mutation Scheduling
Code Coverage
Quantum-behaved Particle Swarm Optimization
Logistic Mapping
Journal
IF:
0.9
Papers:
240
Citations:
1.7K
Organization
No organization information available