Abstract

As optical networks encounter growing traffic demands, efficient bandwidth allocation in Multiband Elastic Optical Networks is essential for optimizing network performance. We examine various bandwidth allocation strategies found in the literature and develop new ones incorporating both heuristic and AI-driven approaches to thoroughly analyze these strategies. Results illustrate the effects of different bandwidth allocation strategies, highlighting the strong performance of methods that utilize spectral band availability to prioritize band selection. We also acknowledge the significance of agent-based solutions, although they necessitate further research to surpass heuristic approaches. © 2025 IFIP TC6 WG6.10.