Hybrid Risk Assessment Model based on Bayesian Networks
François-Xavier Aguessy, Olivier Bettan, Gregory Blanc, Vania Conan and Hervé Debar
Because of the threat posed by advanced multi-step attacks, it is difficult for security operators to fully cover all vulnerabilities when deploying countermeasures. Deploying sensors to monitor attacks exploiting residual vulnerabilities is not sufficient and new tools are needed to assess the risk associated with the security events produced by these sensors. Although attack graphs were proposed to represent known multi-step attacks occurring in an information system, they are not directly suited for dynamic risk assessment.
In this paper, we present the Hybrid Risk Assessment Model (HRAM), a Bayesian network-based extension to topological attack graphs, capable of handling topological cycles, making it fit for any information system. This hybrid model is subdivided in two complementary models: (1) Dynamic Risk Correlation Models, correlating a chain of alerts with the knowledge on the system to analyse ongoing attacks and provide the hosts' compromise probabilities, and (2) Future Risk Assessment Models, taking into account existing vulnerabilities and current attack status to assess the most likely future attacks. We validate the performance and accuracy of this model on simulated network topologies and against diverse attack scenarios of realistic size.