单模光纤中偏振模色散的仿真模型

利用琼斯矩阵法研究了长单模光纤中偏振模色散的仿真模型.考虑到偏振模色散的随机性,该模型中单模光纤被看作是一系列短双折射光纤段的级联,相邻两段之间耦合角是随机的.研究结果表明,当短双折射光纤段等长时,偏振模色散呈现随波长周期性变化的特点;不符合实际情况.当短双折射光纤段不等长且服从高斯分布时,周期性逐渐消失;当其长度均方差为均值的20%,周期性完全消失.最后比较了偏振模色散的时域统计特性.取短双折射光纤段的长度服从高斯分布且均方差为均值的20%,偏振模色散的统计特性接近于实际分布.因此得出结论:为了正确估计偏振模色散的影响,在单模光纤的级联模型中,短双折射光纤段的长度应服从高斯分布,均方差为其均值的20%.

A Simulation Model for Polarization Mode Dispersion in Long Single Mode Fibers

To evaluate the impact of polarization mode dispersion, a long single mode fiber is modeled as a cascade of short birefringence sections with random coupling angles between the adjacent ones (namely cascaded model). The polarization mode dispersion is analyzed by the Jones matrix method. If the individual birefringence sections are of the same length, polarization mode dispersion varies periodically with the wavelength. It is far from the practical case. However, the wavelength periodicity elapses gradually when the individual sections length follows Gaussian distribution. With the standard deviation increasing up to 20% of the mean length, the periodicity disappears completely. Finally, the statistics of polarization mode dispersion is researched. It is concluded that, in the cascaded model for a long single mode fiber, the length of the short sections should follow Gausssian distribution with the standard deviation 20% of the mean value.