Game Theoretic Analysis of Workload Factoring in Federation of Clouds

The customers of the SaaS are distributed all over the world and their QoS may be different from each other too. The SaaS Providers will rent resources from many different IaaS providers or many different data centers in different geographic locations to provide excellent services and meet different QoS of different customers. In this paper, we present a workload factoring model in federation of clouds based on game theory, in which each SaaS provider is considered as a selfish player attempting to split its load between multiple clouds and local resources to minimize the completion time of its job. We demonstrate that the considered game admits a Nash equilibrium by analyzing the selfish players´ behavior. At last, we give an algorithm to find the Nash equilibrium. The obtained results provide the guidelines for user to use these resources more efficient.