A Vulnerability and Exploit Independent Approach for Attack Path Prediction

Network Security has gained an ever increasing importance today because of the increased dependence on the networks. One way to assess the threats to the networks is through the use of the attack graphs. However, because of their sheer enormity and complexity the analysis becomes difficult. Given an attack graph the identification of a probable attack path still remains an issue. This paper presents an attack path identification methodology which is both efficient and scalable. The proposed approach is based on the attack surface measure of the individual hosts comprising the network and the access levels between them and it identifies the attack path independent of the vulnerabilities or the exploits that may exist.