Off-Peak Energy Optimization for Links in Virtualized Network Environment

Energy consumption in information and communication technology (ICT) is estimated to be 10 percent of the total energy consumed in industrial countries. Besides, the population of ICT customers is growing. In order to handle the increasing traffic demands, service providers need to expand their network infrastructure. The recent proposed network virtualization technology helps slow down the infrastructure expansion by allowing the coexistence of multiple virtual networks over a single physical network. Although virtualized network environment (VNE) slows down the infrastructure expansion and therefore controls power consumption, it is essential to develop new techniques to decrease VNE's energy consumption. In this paper, we discuss multiple novel energy saving reconfiguration methods that globally/locally optimize VNE's link power consumption, during off-peak time. The proposed fine-grained local reconfiguration enables the providers to adjust level of the reconfiguration, and accordingly control possible traffic disruptions. An Integer Linear Program (ILP) is formulated for each solution according to two power models, and considering the impact of traffic splitability. Because the formulated ILPs are not scalable to large network sizes, a novel heuristic algorithm is also suggested. The simulation results prove the proposed solutions are able to save notable amount of energy in physical links during off-peak time.