Predicting strategic behavior using game theory for secure virtual machine allocation in cloud

Abstract

In cloud computing, security is a very crucial issue to be taken care of. So, game theory can be used as a strategic tool in decision making among users to protect their virtual machines that are hosted on same hypervisor from external as well as internal attacks. In this paper, a payoff matrix and a decision tree are constructed for ‘n’ number of users, and problem of interdependency is dealt with. Then, a unique user is chosen and his choices of investing in security are evaluated till equilibrium is achieved, when attacker launches his ‘m’ number of strategies. Finally, payoff values of each are calculated, and a decision tree is built which shows the strategies for both attacker and jth user, and Nash Equilibrium is also calculated which shows the best response for both players against each other.