Capacity-aware resource provisioning by prioritizing the highest capacity band in next-generation multi-band elastic optical networks

Abstract

The emergence of next-generation multi-band elastic optical networks (MB-EONs) marks a paradigm shift in addressing the escalating bandwidth demands of future communication systems. While conventional C-band provisioning is nearing its spectral saturation point, expanding into additional spectral bands such as L, S, and E bands provides an avenue for scalable capacity enhancement. However, utilization of these heterogeneous spectral resources necessitates efficient provisioning strategies. This paper proposes a capacity aware-resource provisioning technique that prioritizes spectral bands based on their spectral capacity, favoring the highest-capacity band during allocation unlike traditional provisioning methods in the literature which prioritize the C-band for resource provisioning which has least resource availability. Through comprehensive simulation-based evaluation, we show that band sequencing strategies can substantially affect the network performance. Simulations have been conducted on NSF and USNET topologies by considering the following performance metrics: request blocking probability (RBP), bandwidth blocking probability (BBP), spectral efficiency (SE), total bits transmitted, band utilization distribution (BUD) and band resource allocation time. Simulation results demonstrate that our strategy significantly outperforms baseline models in terms of blocking probabilities with a least reduction of RBP and BBP by 0.6% and 2.4% respectively.