Game Analysis of Workload Factoring with the Hybrid Cloud

Cloud computing is a novel scalable and on-demand computing service where computing resources are offered over the Internet. Under the situation of coexistence of enterprise local resources and public cloud resources, the ultimate goal of in-house users is increasing their revenues by making the right mix decision for resources utilization. This paper introduces a free workload competing game model in which in-house users consider utility influenced by two primary factors: process cost and completion time. Firstly, for analyzing generalized utility functions of users, the cost-time indifference utility curve is proposed to describe the relationship of process cost, completion time and utility of the job. Secondly, we construct a general non-cooperative game model for resources utilization in which the generalized utility functions are classified into three types of indifference utility curve. By theoretical analysis, there must exist at least one Nash equilibrium for straight line and convex to origin of indifference utility curve, and the equilibrium exists when the utility function of each user is identical and homogeneity for concave to origin. Furthermore, we propose two suggestions for enterprise making right mix decisions for constructing and utilizing of private cloud based on the analysis of user´s actions in hybrid cloud.