Optimal Virtual Machine Provisioning in Cloud Computing Using Game Theory

Virtualization of resources on the cloud computing allow fast and easy scaling of infrastructure to achieve a specified QoS. This will undoubtedly improve the allocation efficiency of physical machines in large-scale data centers. In fact, virtualization technologies not only have a direct effect on the performance and energy consumption but also can reduce the operating costs considerably. As there is fierce competition among cloud providers, there is an urgent need to adopt an elastic resource management system and an optimal pricing strategy as well. Such a strategy is designed principally to satisfy the requirements of a given service and simultaneously increase its associated profit margins. This study highlights why game theory is a very useful tool to analyze the impact of both energy efficiency and response time on virtual machine (VM) provisioning. To this aim, we present a formal policy that ensures fairness in the resource sharing and also takes into account the reasonable expectations market participants. More precisely, we rely on a non-cooperative game model in choosing the right provisioning scheme, and then discuss its equilibrium and stability. In a duopoly market, we conduct some numerical experiments to illustrate the efficiency and performance of the proposed approach and its proof of convergence with a certain number of iterations. © 2021, Springer Nature Switzerland AG.