Low-Complexity and Adaptive Nonlinearity Estimation Module Based on Godard's Error

Abstract

A low-complexity and adaptive nonlinearity estimation module is proposed and experimentally demonstrated based on Godard's error and low-pass filter (LPF). The computational complexity of the proposed adaptive estimation module is significantly reduced by avoiding utilizing the polarization demultiplexing, frequency offset compensation, and carrier phase recovery in the process of finding the optimal value $\gamma\xi_{\mathrm{opt}}$ compared with the previous approach based on phase noise variance. The performance of the proposed module is experimentally verified in a 40-Gb/s polarization-division-multiplexing-quadrature-amplitude modulation (PDM-QPSK) coherent optical communication system over 720-km single-mode fiber (SMF). Experimental results show that the computational complexity of the proposed nonlinearity estimation module is only ∼7.35% of the previous approach and ∼27.8% of that with an LPF inserted.