A Review of Multi Agent Perimeter Defense Games

Abstract

This paper reviews a series of works done on the multi-agent perimeter defense scenario, in which a team of intruders try to score by reaching the target region while a team of defenders try to minimize the score by intercepting those intruders. We describe how the small-scale differential games are solved and are leveraged to design team strategies in the large-scale swarm versus swarm scenarios. Three different approaches to analyze the large-scale games (MM, MIS, and LGR) are introduced with comments on their relative strengths and weaknesses. As a unique contribution of this paper, we discuss how the existing results can be extended into more general problem formulations. Furthermore, we point out the limitations of the current work and suggest potential directions for future research.