On the Efficiency of Importance Sampling Techniques for Polarization-Mode Dispersion in Optical Fiber Transmission Systems

An adaptive variance reduction technique was recently proposed to efficiently study transmission errors induced by polarization-mode dispersion (PMD) in optical fiber transmission systems. The technique combines importance sampling (IS) and the cross entropy (CE) method to bias Monte Carlo simulations. Here we (i) present a detailed study to quantify the computational efficiency of this technique, (ii) introduce a novel variant of the hinge model for PMD in optical fiber communication systems, and (iii) use the IS-CE technique to compare the statistical properties of various hinge models. Specifically, we compare the statistics of PMD-induced transmission impairments due to the traditional hinge model (isotropic hinge output), the waveplate hinge model (anisotropic hinge output), and the novel variant of the waveplate hinge model introduced here. We use computationally efficient expressions to determine the probability density function of the differential group delay, and we couple these techniques with the outage map method to compute statistical probabilities of system outages, as quantified by the noncompliant capacity ratio.