Energy-saving traffic scheduling in backbone networks with software-defined networks

Abstract

The rapid development of communications networks facilitate our lives but bring severe energy consumption problems. As an emerging network structure, software-defined networking can realize flexible scheduling of network traffic demands. In this paper, we exploit the above feature and adopt the idea that is powering down unnecessary links and improving links bandwidth’s utilization to formulate an optimization problem. It can realize reducing network costs while satisfying traffic demands and network resources. We propose a heuristic algorithm based on the maximum bandwidth utilization of activated links to solve this optimization problem. It aims to minimize energy costs by improving link utilization and turning on network devices as few as possible. Besides, we propose to divide all networks in a library of test instances for Survivable fixed telecommunication Network Design into four categories according to the network complexity. Finally, we choose eight different sizes of networks to verify our algorithm. The results show that the proposed algorithm can realize energy-efficient and improve the energy-saving rate by 0.34–19.69% on average by improving the utilization of active links.