Statistical analysis of pulse propagation driven by polarization-mode dispersion

Josselin Garnier, Julien Fatome, G. Le MeurJournal of the Optical Society of America B 2002 · DOI

TL;DR

This paper provides a comprehensive statistical analysis of how optical pulses behave when affected by random polarization-mode dispersion (PMD) in optical fibers. The authors derive analytical formulas for key pulse characteristics like width, timing, and polarization degree, offering crucial insights for designing more robust fiber optic communication systems.

What Problem Does This Paper Address?

Optical fibers, intended to carry light signals, are not perfect. Imperfections in the fiber's structure lead to birefringence, meaning light traveling in different polarization states experiences slightly different speeds. This phenomenon, known as Polarization Mode Dispersion (PMD), causes a single input pulse to split into two, arriving at different times and broadening the overall pulse. While PMD has been studied, understanding its statistical behavior, especially how it affects pulse characteristics over long distances and for broadband pulses where frequency dependence becomes significant, remains a challenge. This paper aims to fill this gap by providing analytical tools to predict and understand these statistical effects.

Methods

The authors employ a theoretical framework based on stochastic differential equations and the theory of moments. They model the random birefringence as a continuous process described by the Stokes vector and Pauli matrices. Key analytical techniques include:

  • Stokes Space Representation: The evolution of the optical field is analyzed in Stokes space, which provides a geometric representation of polarization states.
  • Stochastic Differential Equations (SDEs): The random nature of PMD is captured using SDEs, allowing for the derivation of statistical properties.
  • Infinitesimal Generators and Fokker-Planck Equations: These are used to derive the probability density functions (PDFs) of various pulse parameters.
  • Analytical Derivations: The paper focuses on deriving closed-form analytical expressions for the mean, variance, and PDFs of pulse width, timing displacement, and degree of polarization.
  • Numerical Simulations: The theoretical results are validated against numerical simulations to confirm their accuracy.

Datasets

Not discussed in the paper. The study relies on theoretical models and numerical simulations rather than experimental datasets.

Key Insights and Ideas

  • Frequency Correlation is Crucial: The paper highlights that the frequency correlation between different polarization modes plays a significant role in how PMD affects pulse characteristics, especially for broadband pulses.
  • Universal Pulse Broadening: The square of the pulse width (T²w) shows a universal linear growth with propagation distance, independent of the initial pulse shape. This is a remarkable finding for understanding long-term PMD effects.
  • Gaussian Approximation for Timing: While not strictly Gaussian, the timing displacement (Tc) can be well-approximated by a Gaussian distribution for practical purposes, especially for shorter fiber lengths.
  • Decay of Polarization Coherence: The degree of polarization of a pulse decays over distance, indicating a loss of polarization coherence among its spectral components due to PMD.
  • GVD's Limited Impact on Timing: Group Velocity Dispersion (GVD) has a negligible impact on the timing displacement of a pulse, but it does enhance pulse broadening.

Key Results

The paper provides analytical expressions for:

  • Mean, Variance, and PDF of Time Displacement (Tc): For a Gaussian pulse, the variance of Tc is given by (T²w) = (T²₀) (1 + 4σ²z/T²₀).
  • Mean, Variance, and PDF of Pulse Width (Tw): The mean square pulse width shows linear growth: (T²w) = T²₀ + 3σ²z. The variance also shows a dependence on PMD strength and distance.
  • Mean, Variance, and PDF of Degree of Polarization (Pd): The degree of polarization decays linearly with distance initially and then approaches zero for long propagation distances.
  • Influence of GVD: GVD enhances pulse broadening but does not significantly affect timing displacement.

Limitations and Future Work

  • Idealized PMD Model: The paper primarily uses a simplified model of random birefringence. Real fibers might exhibit more complex birefringence patterns.
  • Polarization-Dependent Losses (PDL): The study does not explicitly account for polarization-dependent losses, which can also impact signal quality. The authors suggest this as a potential area for future work.
  • Nonlinear Effects: The analysis is limited to linear propagation. Future work could explore the interplay of PMD with nonlinear effects in optical fibers.
  • Experimental Validation: While numerical simulations are used, direct experimental validation of all derived formulas would strengthen the findings.

Summary updated

Abstract

The linear propagation of pulses driven by random polarization-mode dispersion is considered. Analytical expressions are derived for the probability-density functions of the pulse width, timing displacement, and degree of polarization. The study is performed in Stokes space, and frequency correlation between modes is shown to play an important role in it.
BibTeX
@article{garnier2002statistical,
  author = {Josselin Garnier and Julien Fatome and G. Le Meur},
  title = {{Statistical analysis of pulse propagation driven by polarization-mode dispersion}},
  journal = {Journal of the Optical Society of America B},
  year = {2002},
  doi = {10.1364/josab.19.001968},
}
Figure