Optimizing deployment of VMs in cloud computing environment

IaaS is one of the basic service models in cloud computing, and virtual machine deployment is an important task in IaaS. In this paper, we study the problem of data transfer cost between VMs and data centers in cloud computing environment with multi clusters. We propose a measurement model which defines the traffic matrix and unit price matrix, while the optimization target is the total cost of data transfer. Based on the model, we use Hill-climbing Algorithm to get the optimized VM deployment solution. The simulation results show that the traffic cost generated by the Hill-climbing optimization approach is decreased significantly compared with the random deployment method.