Protecting Web Applications from Application Level Security Risks

Abstract

Application level vulnerabilities have become a serious issue to web applications due to their universal exposure.Signature-based detection techniques increasingly fail to cope with new attack types. As a remedy, this thesis presents an effective approach for detecting web application security risks using Web Application Firewall architecture. It comprises of a positive security model that profiles web application behaviour and rejects anomalies along with a negative security model consisting of strong attack signatures. In addition to the existing attack signature database the administrator will have the ability to add new attack signatures. As opposed to existing approaches, the Web Application Firewall architecture does not require modification of the application source code and is capable of detecting attacks that network firewalls and intrusion detection systems fail in.