Enhancing Security of Web-Based IoT Services via XSS Vulnerability Detection

Abstract

The Internet of Things (IoT) technology is experiencing significant growth and integration into various aspects of daily life. With the rising number of connected devices, diverse security challenges are emerging as substantial threats to IoT. Cross-Site Scripting (XSS) is one of the major security risks in web services and so is within the application layer of IoT. Many existing web applications remain susceptible to XSS vulnerabilities. In this paper, we propose an XSS detection scheme aimed at enhancing the security of IoT, particularly concerning web application services. To achieve this, we developed a framework for combining symbolic execution and dynamic taint analysis to provide a comprehensive security assessment. Our objective is to increase the ratio of vulnerability detection while avoiding false alarms and keeping the required analysis time as minimal. To realize our idea, we have defined an instrumentation scheme for taint analysis and concolic executions and automated the process of vulnerability detection for a web application. Our framework is capable of pinpointing the precise locations of security vulnerabilities and the exact input datasets at risk of XSS threats. Subsequently, the detected flaws can be easily removed. The experimental results demonstrate the validity of the proposed scheme. We achieved a detection rate of XSS threats of 90.62% using a test set of SecuriBench Micro and 69.11% using OWASP while showing 0% false positives.