Game Theoretic Approach for Real-time Task Scheduling in Cloud Computing Environment

Abstract

Cloud computing is one of the emerging technology in the field of distributed computing and is designed as per the requirement and demand of the user. It uses the virtualization technique to create multiple virtual machines that are the basis of computation in the cloud. One of the major issues in cloud computing is to efficiently schedule the tasks and completing their execution before the deadline to maximize the utilization of the processor, maximize the throughput and reduce the waiting time of the task. In this paper, first, we propose a system model and proposed a game-theoretic framework for scheduling real-time tasks in the cloud computing environment to reduce the total completion time and total waiting time. In our game model task act as a player, the virtual machine act as a strategy and the payoff of the player is represented by completion time and waiting time. We have performed our experiments using the non-cooperative and cooperative game model. Our experimental results show that the total execution time and total waiting time are less in the cooperative game model than a non-cooperative game model.