An optimal VM resource allocation for near-client-datacenter for multimedia cloud

With the development of cloud computing, datacenter already plays increasingly important role in improving the multimedia user Quality of Experience (QoE). The resource allocation strategy, namely the virtual machines (VMs) allocation strategy, is a major research direction for cloud computing technology. In this paper, we introduce the definition of near-client-datacenter for virtual machines, which is the main environment for multimedia cloud. Specifically we consider a network environment model for the multi-datacenter. Base on this model, we analyze the key factors in the transmission network between datacenters and propose a resource allocation strategy for the VMs in datacenters to minimize the resource cost. Since the constrained optimization problem is a NP-hard problem, we present a heuristic algorithm to get an approximate solution. The evaluation results not only demonstrate the importance of the transmission network between datacenters in the resource allocation of cloud network, but also show the effectiveness of our optimal heuristic algorithm.