Integrated Vibration Sensing in DSCM Systems Under ECLs Based on ANN and Digital Twin

This paper proposes an integrated vibration sensing and communication scheme for DSCM systems using commercial external cavity lasers (ECLs), enhanced by artificial neural networks (ANN) and digital twin technology. To overcome the limitations of conventional band-pass filtering (BPF) methods in broadband vibration detection, we develop an ANN-based vibration-induced phase extraction scheme. The ANN used for vibration-induced phase extraction is trained with the synthetically generated data from digital twin simulations. The digital twin framework combines experimental noise (captured from vibration-free scenarios) with digitally-added vibration-induced phases to generate training datasets for the ANN training. Experimental validation demonstrates a 12 dB improvement in sensing signal-to-noise ratio (SSNR) compared to traditional BPF-based schemes. This work provides a practical solution for intelligent optical network maintenance by enabling robust vibration sensing without requiring narrow-linewidth lasers or customized hardware.