Modeling and Mitigation of Power Degradation Induced by Hard Decision Errors in Power Profile Estimation

The use of pre-forward error correction (pre-FEC) hard-decision (HD) data is preferred for the practical implementation of the emerging optical performance monitoring technique, power profile estimation (PPE). However, pre-FEC HD data introduces power degradation that increases with the pre-FEC bit error rate (BER). To analyze this effect and mitigate the resulting degradation, we first derive analytical models for HD-induced power degradation in both correlation method (CM)-based and least-squares (LS)-based PPE. With the aid of the Gaussian noise (GN) model and numerical simulations, we show that HD data has a negligible impact on the spatial resolution function (SRF) but introduces significant errors in the accumulated nonlinear vectors. Based on the derived analytical models, we further propose a power degradation mitigation scheme by adjusting the weights of the accumulated nonlinear noise intensity distribution. The effectiveness of the proposed scheme is validated through both 130-GBaud simulations and DP-16QAM 64-GBaud experiments. The results demonstrate that the proposed method achieves estimation accuracy comparable to the ideal post-FEC case for both CM-based and LS-based PPE, with the root mean square error (RMSE) reduced to below 0.5 dB for LS-based PPE, whereas using HD data alone leads to severe degradation. These results support the use of HD data in practical PPE implementations and facilitate its deployment in real-time optical network monitoring.