A Traffic and Resource-aware Energy-Saving Mechanism in Software Defined Networks

Energy-saving mechanisms are crucial in reducing the environmental impact and energy costs of ICT network infrastructures. Integration of renewable energy resources to ICT infrastructures and dynamic energy pricing suggest that not only traffic demands and communication resources but also energy resources should be taken into account in designing energy-saving mechanisms for network infrastructures. With the introduction of Software Defined Networking (SDN) paradigm with a logically centralized controller architecture and programmability of network elements new opportunities have emerged for improving the energy efficiency of the networks. In this paper, by exploiting the SDN features and adopting the idea of powering down unnecessary links to save energy, we formulate an optimization problem for identifying the optimum set of active links that reduces the energy cost of the network while satisfying traffic demands and respecting communication and energy resources. To solve this optimization problem we introduce a computationally efficient heuristic algorithm. Using simulation results, we show that the solutions of the heuristic algorithm reduce the energy cost of the network up to 34% while satisfying the constraints of the problem.