Wavelength defragmentation for seamless service migration

Emerging applications require future elastic optical networks to provide dynamic and reconfigurable services while maintaining efficient usage of wavelength resources. In this paper, we propose integer linear programming (ILP)-based wavelength defragmentation solutions with minimal optical path disruptions during the migration process. The new approaches achieve up to 20% resource efficiency improvement with less than a quarter of disruptions compared to a conventional approach in the sample networks. In addition, we enhance our approach to shorten the computation time and evaluate it under a computation time limitation of 10 min. The enhanced approach also achieves a resource efficiency improvement of more than 20% with no optical path disruption in the realistic networks, compared with the case where defragmentation is not supported.