Bit rate and path length–aware spectrum allocation sequences for multiband elastic optical networks

As multiband elastic optical networks (MB-EONs) evolve, managing spectrum resources efficiently becomes fundamental to meeting growing traffic demands, where heterogeneous bandwidth demands must be provisioned across the C, L, S, and E spectral bands. Existing multiband provisioning approaches typically adopt fixed spectrum allocation sequences, which limit adaptability under varying traffic patterns and network conditions. To address this, we propose three adaptive spectrum allocation strategies: Bit Rate–Aware Spectrum Allocation Sequence (BR-SAS), Path Length–Aware Spectrum Allocation Sequence (PL-SAS), and a joint BR + PL-SAS method. These algorithms dynamically select between the C → L → S → E and E → S → L → C allocation sequences based on the bit rate and path length of incoming requests. Extensive simulations on the NSFNET and USNET topologies evaluate the proposed strategies in terms of request blocking probability (RBP), fairness, and band-sequence utilization behavior. Results show that among all three proposed algorithms, BR + PL-SAS achieves the lowest RBP across both topologies, and BR-SAS provides moderate performance in both RBP and fairness. In contrast, PL-SAS yields the highest fairness due to its balanced band utilization but exhibits the highest RBP, indicating reduced provisioning efficiency. Overall, the proposed BR + PL-SAS algorithm emerges as the most effective solution, outperforming single-parameter strategies and demonstrating the benefits of jointly considering bit rate and path length characteristics in multiband spectrum allocation.