Research On And Implementation Of XSS Vulnerability Mining Based On Fuzzing Technology

With the rapid development of Internet, the functionality and interaction of Web applications are increasing continuously. A user is not only the viewer of Web contents, but also the creator. Due to the diversity of sources of information in a website, the current attacks against Web application vulnerabilities are more than two-thirds of all attacks against security vulnerabilities. Among all attacks on Internet, the cross-site scripting (cross-site scripting, XSS) attacks have become a most popular attack method. XSS attackers steal private information of users by injecting malicious script code in web pages which can operate automatically and secretly in browser while someone browsing the web pages. How to detect potential XSS vulnerability in website is the hotspot of current researches on vulnerability detection technology.Current vulnerability detection technologies still await perfection because of some shortcomings such as needs of open source, high false positive rate and false negative rate and limitation for only known vulnerabilities, so that further research is necessary. The principles, implementation techniques and current researches of XSS vulnerability detection are deeply studied in this paper. The main works of this paper are as follows:Firstly, a novel XSS vulnerability detection model is proposed based on Fuzzing technology, which has advantages of ease to implement, no source restrictions and high automation. Secondly, a dynamic generation algorithm of test data for context-free grammar is designed based on filtering and deleting mechanism of servers aiming at the problem that unknown vulnerabilities can not be detected by current technology. The algorithm helps to detect unknown vulnerabilities by generate new test data instead of pre-defined ones. Thirdly, a real-time detection scheme for target website combining the techniques of HTML parser and API hook is proposed, which automates the detection system.A prototype system of XSS vulnerability detection based on Fuzzing technology is implemented finally using PERL language. Experimental results show that our prototype system can make an exhaustive traversal search on target website and effectively discover the existing XSS vulnerabilities in website. This prototype system is proved to be intelligent and automatic.