Geometrical Pruning of the First Order Regular Perturbation Kernels of the Manakov Equation

Abstract

We propose an approach for constraining the set of nonlinear coefficients of the conventional first-order regular perturbation (FRP) model of the Manakov Equation. We identify the largest contributions in the FRP model and provide geometrical insights into the distribution of their magnitudes in a three-dimensional space. As a result, a multi-plane hyperbolic constraint is introduced. A closed-form upper bound on the constrained set of nonlinear coefficients is given. We also report on the performance characterization of the FRP with multi-plane hyperbolic constraint and show that it reduces the overall complexity of the FRP model with minimal penalties in accuracy. For a 120 km standard single-mode fiber transmission, at 60 Gbaud with DP-16QAM, a 93% reduction in modeling complexity with a penalty below 0.1 dB is achieved with respect to FRP $M=15$ .