Computing the equivalent Poisson traffic for the Wavelength Decomposition Method

Abstract

We present a comparison between different methods to calculate the equivalent Poisson traffic step used by the Wavelength Decomposition Method. The Wavelength Decomposition Method is used to compute the blocking probabilities in Wavelength Division Multiplexing (WDM) optical networks without wavelength converters. The approach divides the WDM network into layers (colors) and uses a moment matching method to calculate an equivalent Poisson overflow traffic to each layer. A corresponding single link model is developed to match the traffic characteristic for each end-to-end traffic. Analyzing blocking probabilities in each layer of the network is derived from an exact approach. We study Fredericks and Hayward's, Sanders's, Equivalent Random Traffic ERT, Rapp's, Berkeley's and Bernoulli–Poisson–Pascal BPP moments matching methods. Simulation results are presented to evaluate the accuracy of each method.