Application of multi-objective bee colony optimization algorithm to Automated Red Teaming

Automated Red Teaming (ART) is an automated process for Manual Red Teaming which is a technique frequently used by the Military Operational Analysis community to uncover vulnerabilities in operational tactics. The ART makes use of multi-objective evolutionary algorithms such as SPEAII and NSGAII to effectively find a set of non-dominated solutions from a large search space. This paper investigates the use of a multi-objective bee colony optimization (MOBCO) algorithm with Automated Red Teaming. The performance of the MOBCO algorithm is first compared with a well known evolutionary algorithm NSGAII using a set of benchmark functions. The MOBCO algorithm is then integrated into the ART framework and tested using a maritime case study involving the defence of an anchorage. Our experimental results show that the MOBCO algorithm proposed is able to achieve comparable or better results compared to NSGAII in both the benchmark function and the ART maritime scenario.