Fading suppression method based on redundant data within the spatial resolution and deep learning for a Φ-OTDR...

Area of Science:

• Optical Engineering
• Signal Processing
• Artificial Intelligence

Background:

• Interference fading in phase-sensitive optical time-domain reflectometry (Φ-OTDR) systems degrades sensing performance.
• Existing fading suppression methods often require complex hardware modifications to the light source.

Purpose of the Study:

• To propose a novel multi-channel data synthesizing deep neural network (MDS-DNN) method for reducing interference fading in Φ-OTDR.
• To enhance the signal-to-noise ratio (SNR) and reduce the false alarm rate without altering the conventional Φ-OTDR setup.

Main Methods:

• Utilizing the redundant information from spatially oversampled data in Φ-OTDR systems.
• Developing a long short-term memory (LSTM) network-based framework for end-to-end training.
• Synthesizing multi-channel data to learn correlations with the ideal sensing signal.

Main Results:

• The MDS-DNN algorithm effectively suppresses phase noise and improves SNR at fading positions.
• Experimental results show an output SNR of 49.88 dB, a 19.65 dB increase over input channels.
• The method reduced the false alarm rate caused by interference fading by one order of magnitude.

Conclusions:

• The MDS-DNN method offers an efficient solution to mitigate interference fading in Φ-OTDR.
• This approach enhances sensing performance without requiring modifications to existing Φ-OTDR hardware.
• The deep learning-based method significantly improves SNR and reduces false alarms in practical Φ-OTDR systems.