Adaptive importance sampling for performance evaluation and parameter optimization of communication systems

In this paper, we present new adaptive importance sampling techniques based on stochastic Newton recursions. Their applicability to the performance ev aluation of communication systems is studied, Besides hit-error rate (BER) estimation, the techniques are used for system parameter optimization, Two system models that are analytically tractable are employed to demonstrate t he validity of the techniques. As an application to situations that are ana lytically intractable and numerically intensive, the influence of crosstalk in a wavelength division multiplexing (WDM) crossconnect is assessed. Tn o rder to consider a realistic system model, optimal setting of thresholds in the detector is carried out while estimating error rate performances, Resu lting BER estimates indicate that the tolerable crosstalk levels are signif icantly higher than predicted in the literature. This finding has a strong impact on the design of WDM networks. Power penalties induced by the additi on of channels can also be accurately predicted in short run-times.