Performance-aware virtual machine allocation approach in an intercloud environment

Cloud Computing raises major challenges among which virtual machine (VM) placement is considered as one of the crucial problems. From the cloud provider's perspective, this process often consists in choosing the best convenient physical hosts with regards to energy efficiency, resource utilization and revenues maximization. However, to secure the loyalty of their clients, cloud providers should also consider the applications performance. In this paper, we propose a VM placement approach tackling the applications performance concern, in an intercloud environment. We state the VM placement as a mixed integer programming problem which aims to maximize a global utility function considering VM bandwidth requirements and network latencies. The proposed approach is evaluated with respect to demands and latencies variations; and through numerical results, we have demonstrated that our model allows us to compute the optimal configuration, in terms of high bandwidth availability and low network latency in the overall intercloud environment.