结合静态分析的软件脆弱性测试方法.pdfVIP

计算机工程与设计 Computer Engineering and Design 1698 20 11, Vol.32, No.5 计算机工程与设计 Computer Engineering and Design 结合静态分析的软件脆弱性测试方法 王 丹， 刘 勇， 吴 咏， 朱鲁华 ( 江南计算技术研究所，江苏无锡2 14083) 摘 要：Fuzzing 是一种有效的软件脆弱性动态测试方法，为了避免传统Fuzzing 技术的盲目性所导致的测试效率低下的问 题，提出了静态分析和遗传算法相结合来指导Fuzzing 数据生成的方法，并设计实现了一个智能化的Fuzzing 系统SFS (smart ) fuzzing system 。触发脆弱点的前提是覆盖脆弱性语句，该方法通过静态分析技术提取脆弱性语句，并用遗传算法指导测试用 例对脆弱性语句进行覆盖，从而提高脆弱性语句的覆盖效率，避免Fuzzing 的盲目性。 关键词：Fuzzing; 静态分析; 遗传算法; 脆弱性语句; 语句覆盖 ( ) 中图法分类号：TP311 文献标识码：A 文章编号：1000-7024 20 11 05-1698-04 Software vulnerability testing method with static analysis WANG Dan, LIU Yong, WU Yong, ZHU Lu-hua (Jiangnan Institute of Computing Technology, Wuxi 2 14083, China) Abstract ：Fuzzing is an effective method of software vulnerability dynamic testing. To avoid the low efficiency due to the blindness of traditional Fuzzing technology, a method of guiding Fuzzing data generation by the combination of static analysis and genetic algorithm is presented, and a Fuzzing system SFS (smart Fuzzing system) is designed and implemented. The premise of triggering vulnerable points is the coverage of vulnerable statements. By using static analysis to extract vulnerable statements and using genetic algorithm to cover vul- nerable statements, this method can improve the efficiency of the vulnerable statement coverage and avoid the blindness of Fuzzing. Key words ：Fuzzing; static analysis; genetic algorithm; vulnerable statement; statement coverage 的代码或路径，难以发现复杂脆弱点。 0 引 言 Fuzzing 作为一种高效的软件脆弱性发现技术已经广泛 软件自身的安全性是影响信息安全的主要因素，而在软 应用于各种软件的安全测