Computing Optimal Attack Strategies Using Unconstrained Influence Diagrams
Attack graphs are a formalism for capturing the most important ways to compromise a system. They are used for evaluating risks and designing appropriate countermeasures. Analysis of attack graphs sometimes requires computing the optimal attack strategy that minimizes the expected cost of the attacker in case of stochastically failing actions. We point out several results in AI literature that are highly relevant to this problem, but remain unnoticed by security literature. We note the problem has been shown to be NP-hard and we present how the problem can be reduced to the problem of solving an unconstrained influence diagram (UID). We use an existing UID solver to assess the scalability of the approach, showing that it can be used to optimally solve attack graphs with up to 20 attack actions.
Keywordsdependency attack graph optimal attack strategy unconstrained influence diagram
Unable to display preview. Download preview PDF.
- 2.Sarraute, C., Richarte, G., Lucángeli Obes, J.: An algorithm to find optimal attack paths in nondeterministic scenarios. In: Proceedings of the 4th ACM Workshop on Security and Artificial Intelligence, AISec 2011, pp. 71–80. ACM, New York (2011)Google Scholar
- 4.Jensen, F.V., Vomlelová, M.: Unconstrained influence diagrams. In: Proceedings of the Eighteenth Conference on Uncertainty in Artificial Intelligence, UAI 2002, pp. 234–241. Morgan Kaufmann Publishers Inc., San Francisco (2002)Google Scholar
- 5.Ingols, K., Lippmann, R., Piwowarski, K.: Practical attack graph generation for network defense. In: Proceedings of the 22nd Annual Computer Security Applications Conference, ACSAC 2006, pp. 121–130. IEEE Computer Society, Washington, DC (2006)Google Scholar
- 8.Homer, J., Ou, X., Schmidt, D.: A sound and practical approach to quantifying security risk in enterprise networks. Technical report, Kansas State University, Computing and Information Sciences Department (2009)Google Scholar
- 9.Luque, M., Nielsen, T.D., Jensen, F.V.: An anytime algorithm for evaluating unconstrained influence diagrams. In: Proc. 4th European Workshop on Probabilistic Graphical Models, pp. 177–184 (2008)Google Scholar
- 10.Isa, J., Lisy, V., Reitermanova, Z., Sykora, O.: Unconstrained influence diagram solver: Guido. In: 19th IEEE International Conference on Tools with Artificial Intelligence, ICTAI 2007, vol. 1, pp. 24–27. IEEE Computer Society (2007)Google Scholar
- 11.Iša, J., Reitermanová, Z., Sýkora, O.: On the complexity of general solution dags. In: Proceedings of the 2009 International Conference on Machine Learning and Applications, ICMLA 2009, pp. 673–678. IEEE Computer Society, Washington, DC (2009)Google Scholar