Transparent Carrier Phase Recovery Based on an Artificial Neural Network

Carrier phase recovery (CPR) is crucial in coherent optical communication, ensuring reliable data transmission. This letter proposes a novel artificial neural network (ANN)- based CPR algorithm. It achieves modulation-format transparency and robust phase recovery across a wide range of signal-to-noise ratios (SNRs) at reduced computational cost compared to a two-stage blind phase search (2S-BPS). Simulations and experiments show that the proposed algorithm outperforms 2S-BPS at low SNRs with shaped quadrature amplitude modulation and surpasses universal carrier phase estimation (U-CPE) and principal component phase estimation (PCPE) with shaped formats at high SNRs. While a 3% pilot-only CPR achieves the best performance at low SNR at lowest computational cost, it does so at the cost of reduced throughput at high SNRs.