An Attack Graph Simulation Approach to Cyber Defense Training

With increasing levels of vulnerabilities, exploits, worms, viruses, and the increasing sophistication of the attacker profile, the training of security personnel is essential in the protection of DoD infrastructure and mission critical information assets. Here we propose an attack graph simulation approach to cyber defense training. In the vulnerability assessment area, attack graphs create a graphical structured model to describe the ways in which a system may be compromised. By using network topology based graph attack simulations that are synched with a vulnerability dictionary, students will be able to understand the ways in which computer networks will be attacked, determine the likelihood and impact of these attacks and decide what action to take where the risks are unacceptable. Our attack graph simulation approach captures attackers' techniques and multi-stage decision processes to develop a vulnerability assessment picture for the student. Our approach will be able to display domain expert's decisions both visually as attack maps and textually as reports to guide students towards achieving the same. In Phase I, we will design, analyze and demonstrate the feasibility of an attack graph simulation approach to cyber defense training by developing student, teacher and curriculum models.