Spectrum-space-divided spectrum allocation approaches in software-defined elastic optical networks

Abstract

Recently, the architecture of elastic optical network (EON) has been proposed as a candidate solution to accommodate both huge bandwidth requirements and flexible connections in next generation optical networks. In order to improve the spectrum efficiency, we propose different spectrum-space-divided approaches and develop two integer linear programming (ILP) models and several spectrum-space-divided spectrum allocation approaches with and without dedicated-path protection in software-defined elastic optical networks (SD-EONs). Simulation results show that the ILP models achieve better performance in terms of the number of frequency slots and hop counts than the proposed spectrum-space-divided spectrum allocation approaches with and without dedicated-path protection under the static scenario of connection requests. Furthermore, we apply the spectrum-space-divided spectrum allocation approaches with and without dedicated-path protection to reduce the blocking probability and to improve spectrum efficiency under the dynamic connection requests compared to the traditional first-fit spectrum allocation approach in SD-EONs.