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

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