Digital Twin-based Insertion Loss Estimator for Anomalous Loss Localization and Network Equalization Enhancement

We propose and experimentally validate a novel accurate digital-twin-based span-level insertion losses estimator. This enables detection and localization of anomalous insertion losses; when combined with equalization, 1.3dB SNR margin improvement is demonstrated despite inaccurate physical layer knowledge.