An Agent Based Social Simulation for evaluating security plan

Abstract

In this research, we propose Agent Based Social Simulation for evaluating security plan. In applying game theory for security, some systems are deployed and developed to allocate randomly limited security resource. Traditional evaluation for security plan, such as mathematical analysis, runtime analysis is not enough in practical situation. An effectiveness of the security plan is depending on attacker performance. In our Agent Based model, attackers act with bounded rational and criminology. Then we evaluate security plan in this model and analyze attackers' dynamically action.