Virtual Machine Placement using Imperialist Competitive Algorithm

The main enabling technology for cloud computing is the use of virtual machines. After making the decision of their placement on hosts, they will be set to run. This process is called virtual machine placement. This process has a great importance on energy consumption and resource wastage avoidance. On the other hand, the growing complexity of cloud infrastructure compounds the problem. In this article, the problem of virtual machine placement is transformed to an optimization problem. The goal is minimizing energy consumption and maximizing the profit of placement, simultaneously. A newly emerged optimization method, called Imperialist Competitive Algorithm is applied in this paper. In addition, a unique method for generating new solutions based on already discovered ones, proposed. Finally the success of the proposed algorithm is confirmed by simulation results and its evaluation is compared with GGA and FFD algorithm.