Request Balancing Among Users in Multiple Autonomous Cloud Provider Environments

摘要

As an increasing number of providers are offering cloud service reservations, an increasingly competitive market has emerged. While most of these providers charge users for metered usage based on fixed prices, they cannot make a timely response to demand changes and may risk losing users. Hence, we highlight the importance of an autonomous provider who self-adjusts the price according to the latest demands of the cloud market. Extensive research has shown that this is not a problem for a single provider. However, when it comes to multiple providers, the main challenge is coordinating the interactions between users and users, providers and providers, as well as users and providers. In this paper, we focus on request balancing among users in the multiple-autonomous-provider environment. Since the request strategy and payment of each user are affected by decisions of other users, we take game theory to address this problem and formulate it into a non-cooperative game among multiple users, without sharing specific strategies. Each user tries to maximize its utility, which combines the net profit with time efficiency. The variational inequality theory is employed and proved that there exists a Nash equilibrium solution set for the game. Then, an iterative proximal algorithm is proposed to find the Nash equilibrium solution, and its convergence is proven. Simulation results demonstrate that the proposed scheme makes requests balanced and increases users' utilities, and also shifts demands from peak to off-peak periods.