Spatio-temporal joint oversampling-downsampling technique for ultra-high resolution fiber optic distributed acoustic...

Area of Science:

• Physics
• Optical Engineering
• Signal Processing

Background:

• Coherent fiber distributed acoustic sensing (DAS) systems are susceptible to noise, limiting their performance.
• Existing noise suppression methods may increase system complexity or introduce crosstalk.

Purpose of the Study:

• To propose and validate a spatio-temporal joint oversampling-downsampling technique for noise suppression in coherent fiber DAS.
• To improve the strain resolution and signal-to-noise ratio (SNR) of DAS systems.

Main Methods:

• Spatial oversampling for artificially dense sampling, followed by spatial downsampling using averaging of differential sub-vectors.
• Temporal oversampling to expand noise bandwidth, ensuring correct noise frequency quantization.
• Temporal downsampling of differential phase reconstruction to recover target bandwidth and reduce out-of-band noise.

Main Results:

• Noise floor inversely correlated with spatiotemporal downsampling factors.
• Achieved strain resolution of 2.58pε/√Hz@100Hz-500Hz and 9.47pε/√Hz@10Hz.
• Reduced probability of large-noise channels from 44.32% to 0% and improved dynamic signal SNR by 20.8dB.
• Optimized noise floor 8dB lower than averaging technique without crosstalk.

Conclusions:

• The proposed technique effectively suppresses noise in coherent fiber DAS systems without increasing complexity or introducing crosstalk.
• The method significantly enhances strain resolution and SNR, making DAS highly competitive for high-precision applications like seismic imaging.