Abstract
We describe the application of importance sampling to Monte-Carlo simulations of polarization-mode dispersion (PMD) in optical fibers. The method allows low-probability PMD events to be simulated much more efficiently than with standard Monte-Carlo methods, thus making it possible to accurately determine the behavior in the tails of the distribution functions associated with a number of PMD statistics. The method also allows the effect of PMD on system outage probabilities to be assessed at realistic levels.
© 2003 Optical Society of America
PDF Article