na o-d m

d m ode geom on a thes atio 2011 detrimental for optical communications as they cause se-Y. Li and A. Yariv Vol. 17, No. 11 /November 2000 /J. Opt. Soc. Am. B 1821vere intermodal interferences and lead to pulse broaden-ing and deformation and hence limit the transmission rate of an optical link. Since these polarization-dependent effects originate largely from random fluctuations along the fiber, their statistical properties are of great importance for assessing their influence on propagating signals and for designing compensation schemes. A rigorous statistical theory of PMD in the absence of PDL was pioneered in 1991 by Fos-chini and Poole.1 In their paper, they proposed a sto-chastic fiber model from which various statistics of PMD could be obtained by use of the standard methods of sto-chastical differential equations. Gisin also indepen-dently applied similar methods to solve the dynamical equation of PMD.2 Following these original studies, the statistics of PMD has received wide and continued interest.3–7 These investigations, among others, have es-tablished a satisfactory, if not complete, theoretical un-derstanding of the statistical nature of PMD in the ab-sence of PDL. est for future studies