静态分析
计算机科学
缓冲区溢出
水准点(测量)
脆弱性(计算)
编码(集合论)
测试套件
脆弱性评估
一套
边距(机器学习)
子程序
动态程序分析
数据挖掘
程序分析
测试用例
计算机安全
机器学习
集合(抽象数据类型)
操作系统
程序设计语言
心理弹性
历史
回归分析
考古
心理学
地理
心理治疗师
大地测量学
作者
Younis Iqbal,Muddassar Azam Sindhu,Muhammad Hassan,Muhammad Amir Javed
标识
DOI:10.1109/iccws53234.2021.9703043
摘要
Buffer overflow (BOF) vulnerability is one of the most dangerous security vulnerability which can be exploited by unwanted users. This vulnerability can be detected by both static and dynamic analysis techniques. For dynamic analysis, execution of the program is required in which the behavior of the program according to specifications is checked while in static analysis the source code is analyzed for security vulnerabilities without execution of code. Despite the fact that many open source and commercial security analysis tools employ static and dynamic methods but there is still a margin for improvement in BOF vulnerability detection capability of these tools. We propose an enhancement in Cppcheck tool for statically detecting BOF vulnerability using data flow analysis in C programs. We have used the Juliet Test Suite to test our approach. We selected two best tools cited in the literature for BOF detection (i.e. Frama-C and Splint) to compare the performance and accuracy of our approach. From the experiments, our proposed approach generated Youden Index of 0.45, Frama-C has only 0.1 Youden's score and Splint generated Youden score of -0.47. These results show that our technique performs better as compared to both Frama-C and Splint static analysis tools.
科研通智能强力驱动
Strongly Powered by AbleSci AI