Benchmarking RWA strategies for dynamically controlled optical networks

We present two studies to benchmark Routing and Wavelength Assignment (RWA) strategies which have been developed for dynamically controlled optical networks. In a first study, we compare the results of a dynamic simulation of RWA using the Suggested Vector (SV) object to reduce wavelength conversion to lower and upper bounds for the off-line wavelength assignment problem defined on the same set of requests. In the second study, the number of blocked requests due to both the dynamic routing and a limitation on the number of conversions per node is compared to an off-line RWA approach. Both studies show that the network performance can be boosted by offline optimization. Bearing in mind the dynamic character of the SV method, its performance is nevertheless acceptable.