Share Email Print

Optical Engineering

Performance characterization of multicanonical Monte Carlo method applied to polarization mode dispersion
Author(s): Alexandre Y. Yamamoto; Aurenice M. Oliveira; Ivan T. Lima
Format Member Price Non-Member Price
PDF $20.00 $25.00

Paper Abstract

The numerical accuracy of the results obtained using the multicanonical Monte Carlo (MMC) algorithm is strongly dependent on the choice of the step size, which is the range of the MMC perturbation from one sample to the next. The proper choice of the MMC step size leads to much faster statistical convergence of the algorithm for the calculation of rare events. One relevant application of this method is the calculation of the probability of the bins in the tail of the discretized probability density function of the differential group delay between the principal states of polarization due to polarization mode dispersion. We observed that the optimum MMC performance is strongly correlated with the inflection point of the actual transition rate from one bin to the next. We also observed that the optimum step size does not correspond to any specific value of the acceptance rate of the transitions in MMC. The results of this study can be applied to the improvement of the performance of MMC applied to the calculation of other rare events of interest in optical communications, such as the bit error ratio and pattern dependence in optical fiber systems with coherent receivers.

Paper Details

Date Published: 7 May 2014
PDF: 6 pages
Opt. Eng. 53(5) 056104 doi: 10.1117/1.OE.53.5.056104
Published in: Optical Engineering Volume 53, Issue 5
Show Author Affiliations
Alexandre Y. Yamamoto, North Dakota State Univ. (United States)
Aurenice M. Oliveira, Michigan Technological Univ. (United States)
Ivan T. Lima, North Dakota State Univ. (United States)

© SPIE. Terms of Use
Back to Top