模糊测试变异策略优化方法

Fuzz Testing Mutation Strategy Optimization Methods

基于覆盖引导的模糊测试技术是当前研究的热点,AFL是该领域的代表性工具。文章在AFL基础上进行了改进,将Havoc的变异策略建模为一个多臂赌博机问题,并提出一种结合ε-greedy算法和置信区间上界(Upper Confidence Bound,UCB)算法的动态调整变异策略的方法,实现了工具EnAFL的设计。通过与AFL的对比实验分析得出,EnAFL在代码覆盖率和测试效率方面表现更出色。

Fuzzy testing technology based on coverage guidance is currently a hot research topic,and AFL is a representative tool in this field.The article improves on AFL by modeling Havoc’s mutation strategy as a multi arm gambling machine problem,and proposes a dynamic adjustment mutation strategy method that combines the ε-greedy algorithm and the Upper Confidence Bound(UCB)algorithm,achieving the design of the tool EnAFL.Through comparative experimental analysis with AFL,it is found that EnAFL performs better in terms of code coverage and testing efficiency.