TC: Small:Automata Based String Analysis for Detecting Vulnerabilities in Web Applications

点击翻译按钮获取中文摘要

Web applications contain numerous vulnerabilities that can be exploitedby attackers to gain unauthorized access to confidential information andto manipulate sensitive data. Many of these vulnerabilities are due toinadequate manipulation of string variables. String analysis, a techniquethat captures the string values that a certain variable might hold at aparticular program point, can be used to identify such flaws. In thisproject, novel and precise string analysis techniques will be developedusing an automata-based approach that represents possible values of astring variable at a program point as an automaton. Techniques thatsupport path-sensitivity and that enable precise analysis of loops usingautomata-based widening operations will be developed. Basic string analysistechniques will be extended to a composite analysis where relationships amongstring variables and other types of variables can be automatically discoveredand analyzed. The precision of string analysis plays a central role forobtaining good results with static vulnerability detection tools. Theprecise string analysis techniques developed in this project will enableanalysis of programs that cannot be analyzed with existing techniques. Theresults of these improved string analysis techniques will lead to novelsoftware security solutions and detection of novel types of vulnerabilities.