Strategies for Mitigating Polarization Fading in a DAS with Homodyne Detection and Delayed Self-Mixing

We present novel algorithms to suppress polarization fading in phase-sensitive Distributed Acoustic Sensing (DAS) using homodyne demodulation with delayed self-mixing in a polarization-diversity optical hybrid. Four phase-combining strategies—equal sum, weighted SNR, weighted visibility, and higher SNR—are evaluated for fading suppression and computational efficiency. Experimental results show that, while all strategies reduce polarization fading, the higher SNR method achieves the best performance, enabling accurate detection of weak acoustic signals up to 3 kHz over a 10-km fiber with less computational complexity and better suitability for real-time applications. The other strategies, while being more robust to polarization fading compared to the conventional method, incur slightly higher computational costs with respect to the higher SNR algorithm. In addition, spatiotemporal analyses confirm precise vibration localization, and a comparison with a commercial DAS system demonstrates the method’s effectiveness for quantitative measurements across a wide spectral range.