Linear Regression vs. Deep Learning for Signal Quality Monitoring in Coherent Optical Systems

Error vector magnitude (EVM) is a metric for assessing the quality of m-ary quadrature amplitude modulation (mQAM) signals. Recently proposed deep learning techniques, e.g., feedforward neural networks (FFNNs) -based EVM estimation scheme leverage fast signal monitoring in coherent optical communication systems. Such estimates from histograms (AHs) short sequences captured before carrier phase recovery (CPR). In this work, we explore further complexity reduction by proposing simple linear regression (LR) method. We systematically compare performance method with FFNN-based and demonstrate its capability to infer an AH when format information known advance. perform both simulation experiment show that LR-based achieves comparable accuracy as scheme. The technique can be embedded identification modules provide comprehensive information. Therefore, work paves way design fast-learning parsimony future intelligent OPM enabler.