基于KLEE的软件漏洞测试用例自动生成技术  

Automatic generation of software vulnerability test case based on KLEE

作　　者：黄琦[1,2] 蔡爱华[1,2] 吕慧颖[3] 彭武 王冬海 陈君华[4] 

机构地区：[1]电子科技大学电子科学技术研究院,四川成都611731 [2]中国电子科技集团公司电子科学研究院,北京100041 [3]首都师范大学管理学院,北京100089 [4]云南民族大学云南省高校物联网应用技术重点实验室,云南昆明650500

出　　处：《计算机工程与设计》2016年第6期1515-1519,1525,共6页Computer Engineering and Design

基　　金：云南民族大学云南省高校物联网应用技术重点实验室开放基金项目

摘　　要：为解决KLEE中STP求解器无法对非线性求解约束条件求解的问题,将改进的遗传算法应用于KLEE中自动生成软件漏洞测试用例。通过改进遗传算法,将其应用于求解非线性方程并编程实现求解模块;通过KLEE的改进,让其能够调用遗传算法求解模块与STP求解器共同运行;通过接口的对接,完成KLEE中两个求解模块的并行运行,自动生成软件漏洞的测试用例。对比实验自动生成测试用例数据与原来KLEE的求解数据,对比结果表明了改进的遗传算法可应用于KLEE自动生成测试用例,其具有可行性、正确性、有效性。The STP solver can not deal with the problems of nonlinear constraints,hence,test cases of the software vulnerabilities were generated automatically based on the improved genetic algorithm which applied in the KELL.The improved genetic algorithm was used to solve nonlinear equations,program and realize the solving module.Through the improvement of KLEE,the improved genetic algorithm was able to call genetic algorithm module and STP solver to run together.Two solving modules were operated in parallel and software vulnerabilities test cases were generated automatically in KLEE by means of docking of interface.By comparing the test cases data generated automatically with experimental and the original solving data from KLEE,the present work verified that the improved genetic algorithm can be applied to generate test cases automatically with the characters of feasibility,correctness and effectiveness.

关 键 词：测试用例 软件漏洞挖掘 改进的遗传算法 KLEE 约束求解 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]