Research On Attack Vector In XSS Detection

With the rapid development and wide application of Internet technology,various web security issues have also become increasingly prominent.Especially,cross-site scripting vulnerabilities have become one of the most common web vulnerabilities in web applications because of which have the characteristics of being easy to use,widespread and destructive.Therefore,it is meaningful and significant in theory and application to research on the detection of potential XSS vulnerabilities in web application.Compared to white box detection technology,black box detection technology has been extensively applied in XSS detection becauser of its high efficiency and no requirements for source code.This thesis is based on the black box detection technology for XSS vulnerability detection,lack kinds of XSS attack vectors,incomplete coverage,poor pertinence,and inadequate application methods are common problems in XSS detection technology,in this thesis,the generation,optimization and application methods of XSS attack vectors are discussed in order to solve these problems and finally an optimized XSS attack vector repertory is built.This thesis proposes a method of identify the output point location which based on the optimized XSS attack vector repertory,which can automatically select XSS attack vectors that meet the output point context.Besides,based on the optimized XSS attack vector repertory and the XSS attack vector selection method,a XSS detection system which is called XSS Scanner is designed and implemented in this thesis.Firstly,the key technologies used in XSS vulnerability detection are analyzed.The HTML knowledge needed to construct the XSS attack vector is studied.In addition,the network crawler technology and the crawler strategy in the XSS black box detection are analyzed and compared.The characteristics and advantages of simulated browser technology Selenium Webdriver are described.What's more,the basic syntax and function of the page information search language XPath are expounded.Secondly,the generation method of XSS attack vector is studied and an initial XSS attack vector repertory is built.This thesis uses an output point context-based XSS attack vector modeling method,explores the composition of XSS attack vectors based on the classification of output points location,design XSS attack vector component and operator expressions which formally describes the composition of XSS attack vectors in different output point contexts.Based on operator expressions,an XSS attack vector instantiation method is proposed.Through the establishment of the component factor database and the design of the two-dimensional table of composition factor constraint relations,the instantiation of XSS attack vector composition model is realized and actual XSS attack vectors are generated and an initial XSS attack vector repertory is built.Furthermore,experiments prove the effectiveness of using this method to generate XSS attack vectors.Next,the variation and application of XSS attack vector are studied.The existing mutation rules according to XSS attack vector are summed up.Based on the browser parsing principle,the mutated position of each mutation rule in the XSS attack vector is designed.Restricted mutation operators are designed to mutate initial XSS attack vectors and then the mutation-optimized XSS attack vector repertory is constructed.Based on the constructed XSS attack vector repertory,this thesis proposes an XPath-based automatic recognition method for output point location,which can automatically select XSS attack vectors that meet the output point context.The shortage of the XSS attack vector random selection method in the existing XSS detection is overcome,and the detection of the invalid XSS attack vector is avoided.Then,the design and implementation of XSS vulnerability detection system XSS scanner are studied.The web crawler module,XSS attack vector library module and XSS detection module are designed and implemented respectively.Finally,Test and analysis of the XSS Scanner XSS vulnerability detection system are designed in this thesis.Test cases are designed to test the basic functions and vulnerability detection capabilities of the XSS Scanner XSS vulnerability detection system,the test results show that the XSS vulnerability detection system designed in this thesis can effectively detect XSS vulnerabilities in web applications.