A Scalable and Accurate Hybrid Vulnerability Analysis Framework

SCHEME: AFR PhD

CALL: 2014

DOMAIN: IS - Information and Communication Technologies

FIRST NAME: Julian

LAST NAME: Thomé

INDUSTRY PARTNERSHIP / PPP: No

INDUSTRY / PPP PARTNER:

HOST INSTITUTION: University of Luxembourg

KEYWORDS: Software Security AssuranceVulnerability AnalysisProgram AnalysisSymbolic ExecutionConstraint SolvingMachine Learning

START: 2014-09-01

END: 2018-04-14

WEBSITE: https://www.uni.lu

Submitted Abstract

As the Internet has become an integral part of our everyday life for activities such as e-mail, online-banking, shopping, entertainment, etc., vulnerabilities in Web software arguably have greater impact than vulnerabilities in other types of software. Vulnerabilities in Web applications may lead to serious issues such as disclosure of confidential data, integrity violation, denial of service, loss of commercial confidence/customer trust, and threats to the continuity of business operations. For companies these issues can result in significant financial losses.The most common and serious threats for Web applications include injection vulnerabilities, where malicious input can be “injected” into the program to alter its intended behavior or the one of another system. These vulnerabilities can cause serious damage to a system and its users. For example, an attacker could compromise the systems underlying the application or gain access to a database containing sensitive information.The goal of this thesis is to provide a scalable approach, based on symbolic execution and constraint solving, which aims to effectively find injection vulnerabilities in the server-side code of Java Web applications and which generates no or few false alarms, minimizes false negatives, overcomes the path explosion problem and enables the solving of complex constraints.

This site uses cookies. By continuing to use this site, you agree to the use of cookies for analytics purposes. Find out more in our Privacy Statement