Resource Assignment based on Core-State Value Evaluation to Handle Crosstalk and Spectrum Fragments in SDM Elastic Optical Networks

A novel routing, core and spectrum assignment (RCSA) algorithm based on reinforcement learning is proposed for space division multiplexing elastic optical networks (SDM-EONs). The algorithm introduces two metrics that measure the spare capacity of the network in terms of the inter-core crosstalk in multi-core fibers and the efficiency of spectrum resource utilization. Numerical simulations indicate that it can enhance the network capacity by up to 46.8% and 10.4%, compared with two conventional schemes.