Lightpath Management in SDN-Based Elastic Optical Networks With Power Consumption Considerations

Abstract

Elastic optical networks (EONs) are increasingly used to interconnect data centers. As interdatacenter traffic is expected to continue to grow at high rates for the foreseeable future, it becomes imperative to address the respective growth in power consumption within EONs. In this paper, we take the first steps toward enhancing the energy efficiency of interdatacenter EONs. Specifically, we use traffic prediction techniques for centralized lightpath management that leverages the capabilities of software defined networking (SDN) technology. Our objective is to eliminate unnecessary lightpath termination and re-establishment operations, and in turn decrease switching power consumption within the network. Our evaluation study confirms that the proposed algorithm is effective in achieving substantial savings in power consumption while maintaining a bandwidth blocking ratio at levels comparable to those of earlier algorithms.