2020_JOCN_Constellation_Dataset

Fast signal quality monitoring for coherent communications enabled by CNN-based EVM estimation

To be published on Journal of Optical Communications and Networking (JOCN) Special Issue on Machine Learning Applied to QoT Estimation in Optical Networks

Authors: Yuchuan Fan, Aleksejs Udalcovs, Xiaodan Pang, Carlos Natalino, Marija Furdek, Sergei Popov, Oskars Ozolins

Abstract: We propose a fast and accurate signal quality monitoring scheme that uses convolutional neural networks (CNN) for error vector magnitude (EVM) estimation in coherent optical communications. We build a regression model to extract EVM information from complex signal constellation diagrams using a small number of received symbols. For the additive white Gaussian noise (AWGN) impaired channel, the proposed EVM estimation scheme shows a normalized mean absolute estimation error of 3.7% for quadrature phase shift keying (QPSK), 2.2% for 16-ary quadrature amplitude modulation (16QAM), and 1.1% for 64QAM signals, requiring only 100 symbols per constellation cluster in each observation period. Therefore, it can be used as a low-complexity alternative to conventional bit-error-rate (BER) estimation, enabling solutions for intelligent optical performance monitoring.