Incentive mechanism design for security investment with local exit equilibrium on structured populations

Abstract

The fact that most of cyber security problems are interdependent security problems, in which one agent’s security is affected not only by this agent’s security investment but also by the security decisions of his neighbors, has encouraged more and more people to study security games and security investment on large computer networks. In this paper, we study interdependent security problems by modeling the problems as public goods games on structured populations. We put forward an incentive mechanism named quantized externality mechanism to improve the overall social payoff in evolutionary public goods games of security investment on complex networks. Moreover, we introduce local exit equilibrium into our new mechanism to satisfy voluntary participation principle. Simulation results show that our mechanism can play an important role in increasing the level of security investment on complex networks. We also find that the participation rate of our new mechanism can converge to a stable level.