Relative cost routing and spectrum allocation in elastic optical networks

Elastic optical networks (EONs) are being used to adapt to the recent rapid changes in traffic behavior and the exponential growth in bandwidth demand. Routing and spectrum allocation (RSA) is one of the most important determining factors of the design, performance, and operation of EONs. In this work, we propose an adaptive RSA algorithm based on the relative cost concept under which the transient effect when a call is admitted on a given route/spectrum set at a given network state is estimated. The algorithm computes the relative cost of forwarding an arriving call across all available route/spectrum sets to the destination. The algorithm selects the route/spectrum set with the least relative cost; if the minimum relative cost is more than the call's value, the call is potentially rejected. In this study, the efficiency of using relative cost to solve the RSA problem is demonstrated via simulation. (C) 2020 Optical Society of America