Chromatic dispersion monitoring for coherent optical communications using artificial neural network

Abstract

With the development of coherent optical communication systems, the optical communication networks are becoming more dynamic and complex. The optical signals in communication networks are vulnerable to various impairments. Therefore, optical performance monitoring (OPM) is essential for optical networks, which ensures the robustness of the networks. Real-time chromatic dispersion (CD) monitoring is the basis for the compensation of CD induced transmission impairment. Therefore, to enable the high quality-of-service of reliable optical networks, it is essential to implement accurate CD monitoring for fiber optical transmission link. In this paper, we present a blind CD estimation method in conjunction with artificial neural network (ANN) for polarization division multiplexing (PDM) coherent optical communication systems. The CD monitoring range is set from 1340ps/nm to 9380ps/nm with step of 1340ps/nm. We experimentally demonstrate the proposed method in PDM quadrature phase-shift keying (QPSK), PDM 16 quadrature amplitude modulation (16-QAM) and PDM 64-QAM systems. The experiment results in our scheme show that CD values can be estimated successfully. The root-mean-square errors (RMSEs) of CD monitoring for 112Gbps PDM QPSK, 112Gbps PDM 16-QAM and 120Gbps PDM 64-QAM are 3.79ps/nm, 7.57ps/nm and 9.68ps/nm respectively.