Error estimation in multicanonical Monte Carlo simulations with applications to polarization-mode-dispersion emulators

被引：16

作者：

Lima, AO ^{[1
]}

Lima, IT

Menyuk, CR

机构：

[1] Univ Maryland Baltimore Cty, Dept Comp Sci & Elect Engn, Baltimore, MD 21250 USA

[2] N Dakota State Univ, Dept Elect & Comp Engn, Fargo, ND 58105 USA

来源：

JOURNAL OF LIGHTWAVE TECHNOLOGY | 2005年 / 23卷 / 11期

基金：

美国国家科学基金会;

关键词：

multicanonical Monte Carlo (MMC) simulations; optical communications; polarization-mode dispersion (PMD); statistical error;

D O I：

10.1109/JLT.2005.857728

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

This paper shows how to estimate errors in multi-canonical Monte Carlo (MMC) simulations using a transition-matrix method. MMC is a biasing Monte Carlo technique that allows one to compute the probability of rare events, such as the outage probability in optical-fiber communication systems. Since MMC is a Monte Carlo technique, it is subject to statistical errors, and it is essential to determine their magnitude. Since MMC is a highly nonlinear iterative method, linearized error-propagation techniques and standard error analyses do not work, and a more sophisticated method is needed. The proposed method is based on bootstrap techniques. This method was applied to efficiently estimate the error in the probability density function (pdf) of the differential group delay (DGD) of polarization-mode-dispersion (PMD) emulators that has been calculated using MMC. The method was validated by comparison to the results obtained using a large ensemble of MMC simulations.

引用

页码：3781 / 3789

页数：9