Joint optimization of routing and VM resource allocation for multimedia cloud

摘要

With the development of cloud computing, cloud data centers provide powerful platforms for multimedia services. Still, QoS routing and resource allocation are two main challenges to optimize the multimedia services. The QoS routing affects the transmission time, while the resource allocation determines the resource utilization in cloud. In this paper, we introduce a cloud network model. Based on the model, we analyze the transmission delay among data centers, and propose a QoS-SPA routing algorithm. Furthermore, we study the optimal VM allocation problem with the objective to minimize the resource cost subject to the network transmission delay among data centers. As the optimization problem is NP-hard, we propose an efficient heuristic, price-performance ratio-based heuristic algorithm (PPR-HA), to achieve a sub-optimal solution. The extensive experimental results demonstrate that joint optimization of routing and VM with QoS-SPA and PPR-HA can effectively reduce the transmission delay among data centers and improve the resource utilization in cloud.