A multi-objective genetic algorithm for minimising network security risk and cost

Abstract

Security countermeasures help ensure information security: confidentiality, integrity and availability(CIA), by mitigating possible risks associated with the security event. Due to the fact, that it is often difficult to measure such an impact quantitatively, it is also difficult to deploy appropriate security countermeasures. In this paper, we demonstrate a model of quantitative risk analysis, where an optimisation routine is developed to help a human decision maker to determine the preferred trade-off between investment cost and resulting risk. An offline optimisation routine deploys a genetic algorithm to search for the best countermeasure combination, while multiple risk factors are considered. We conduct an experimentation with real world data, taken from the PTA(Practical Threat Analysis) case study to show that our method is capable of delivering solutions for real world problem data sets. The results show that the multi-objective genetic algorithm (MOGA) approach provides high quality solutions, resulting in better knowledge for decision making.