A Game Theory Approach for Multi-agent System Resources Allocation against Outside Threats

Abstract

Emergency Response Systems (ERS) and Homeland Security Advisory Systems (HSAS) are both centralized Multiagent Systems (MASs) which delegate multiple interacting agents to resist outside attacks. However, the effectiveness of these MASs as a means to defend entire large-scale geographic regions is constrained by the available resources. The system administrator faces a density of agent deployment dilemmas, where the disposition of more agents easily leads to higher costs. These systems also lack specific measures for rational decision-making, and do not apply mathematical models to capture the interactions between attacker and defender. The administrator of a MAS should have a tool to measure the strength of the attacks and the resistance capability of the response agents. By considering the utilities of moves available to the attacker and the defender, we can find a way to build a rating system for decision making [1].