Energy Efficieny in Virtual Machines Allocation for Cloud Data Centers Using the Imperialist Competitive Algorithm

摘要

Considering that the large amount of energy consumption in data centers have been done by physical hosts, reducing the number of physical hosts will have a clear impact on data center's energy consumption. The problem of allocation virtual machines to physical hosts is one of the most important challenges of infrastructure as a service layer in the cloud computing environment and energy consumption can be optimized by minimizing the number of switched on physical hosts. In general, the problem of optimal allocation of virtual machines to physical hosts can be divided into three sub-problems. The first problem asks when should the virtual machine migrate?. The second issue deals with which virtual machine should migrate?. The third issue should response to the question that where should the virtual machine migrate?. The main focus of this study is to solve the third problem. In this paper, a method based on the imperialist competitive algorithm is provided for optimal allocation of virtual machines to physical hosts. In order to evaluate the performance of the proposed approach, a wide range of solutions offered for the first and second issues are combined with the proposed method and their performance are analyzed. The Cloudsim is used for the simulation of the proposed method. The proposed method mitigates the violation of the SLA. Moreover, this approach has caused the energy consumption in physical hosts with 31.25 percent reduction in comparison to similar algorithms.