The Impact of Rayleigh Scattering in UWFBG Array-Based Φ-OTDR and Its Suppression Method

Feng Wang, Yong Yu, Rui Hong, Ruifeng Tian, Yixin Zhang, Xuping ZhangSensors 2023 · OpenReview

TL;DR

This paper investigates how Rayleigh backscattering (RBS) noise impacts the accuracy of ultra-weak fiber Bragg grating (UWFBG) array-based phase-sensitive optical time-domain reflectometry (Φ-OTDR) for distributed acoustic sensing. The authors demonstrate that reducing the probe pulse duration effectively suppresses RBS interference, leading to a significant improvement in measurement precision.

What Problem Does This Paper Address?

Distributed acoustic sensing (DAS) systems are crucial for applications like geological exploration and structural health monitoring. Phase-sensitive optical time-domain reflectometry (Φ-OTDR) is a key technology in DAS, offering high sensitivity and broadband response. While UWFBG arrays have enhanced Φ-OTDR performance by providing a stronger reflected signal compared to traditional Rayleigh backscattering (RBS) methods, RBS can still act as a significant noise source. This noise arises from the interference between the desired reflected signal from the UWFBGs and the weaker, distributed RBS signal along the fiber. This interference can distort the phase information, leading to inaccurate measurements of acoustic vibrations, especially in disturbed fiber sections. The paper aims to quantify this impact and propose a practical method to mitigate it.

Methods

The researchers employed a Φ-OTDR system utilizing a UWFBG array. Their core methodology involved:

  • Theoretical Modeling: They derived equations to describe the interference between the reflected light from UWFBGs and the distributed RBS signal. This model highlights how the superposition of these signals affects the overall phase measurement.
  • Simulation: The theoretical model was used to simulate the beat signal and phase demodulation under different noise conditions (white noise and RBS interference).
  • Experimental Validation: They conducted experiments using a 1.5 km UWFBG array. The key experimental manipulation was varying the duration of the probe pulses. They compared the system's performance with longer pulse durations (e.g., 450 ns and 300 ns) against shorter ones (e.g., 200 ns and 100 ns).
  • Vibration Testing: To assess the impact on actual sensing, they applied controlled sinusoidal vibrations using a piezoelectric transducer (PZT) and analyzed the demodulated phase and amplitude.
  • IQ Demodulation: The acquired time-domain signals were processed using IQ demodulation to extract phase information.

The novelty lies in systematically analyzing the specific impact of RBS interference in the context of UWFBG-based Φ-OTDR and proposing pulse duration reduction as a direct countermeasure.

Datasets

Not discussed in the paper. The experiments were conducted using a custom-built experimental setup with a 1.5 km UWFBG array.

Key Insights and Ideas

  • RBS is a primary noise source: Despite the advantages of UWFBGs, Rayleigh backscattering remains a significant noise contributor in these systems, impacting the accuracy of vibration measurements.
  • Interference is position-dependent: The phase deviation caused by RBS interference is not uniform along the fiber. It is particularly pronounced in disturbed sections where the RBS signal can significantly overlap with the reflected UWFBG signal.
  • Pulse duration is a critical parameter: The duration of the probe pulse directly influences the extent of overlap between the reflected UWFBG signal and the RBS signal. Shorter pulses reduce this overlap.
  • Pulse duration reduction improves precision: Shortening the pulse duration effectively suppresses the contribution of RBS to the beat signal, leading to more accurate phase demodulation and thus more precise reconstruction of vibration amplitudes.

Key Results

  • Qualitative demonstration of RBS impact: Figure 4 clearly shows that as the pulse duration of one of the probe pulses decreases, the contribution of the "reflection-scattering beat" (RBS + reflected light) becomes less dominant compared to the "reflection-reflection beat" (desired signal).
  • Quantitative improvement in accuracy: Experiments showed that reducing the pulse duration from 300 ns to 100 ns reduced the amplitude fluctuation of the recovered vibration signal by approximately one-third. Specifically, the maximum deviation dropped from 0.19 rad to 0.03 rad within 1 hour of measurement.
  • Improved measurement precision: Using a 100 ns pulse duration improved measurement precision by three times compared to a 300 ns pulse duration.

Limitations and Future Work

  • Hardware limitations: The paper acknowledges that reducing pulse duration is limited by the response speed of components like acousto-optic modulators (AOMs). Insufficient response speed can lead to amplitude fluctuations in the beat signal.
  • Demodulation requirements: The phase demodulation process requires sufficient effective signals, and a very short pulse might not provide enough signal for robust low-pass filtering, especially in I/Q demodulation.
  • Future work: The authors suggest that the selection of pulse duration needs to balance the need for accuracy (shorter pulses) with hardware capabilities and demodulation requirements. Future work could explore optimized pulse shaping or advanced signal processing techniques to further mitigate RBS effects while maintaining high response speeds.

Summary updated

Abstract

Ultra-weak fiber Bragg grating (UWFBG) array-based phase-sensitive optical time-domain reflectometry (Φ-OTDR) utilizes the interference interaction between the reference light and the reflected light from the broadband gratings for sensing. It significantly improves the performance of the distributed acoustic sensing system (DAS) because the intensity of the reflected signal is much higher than that of the Rayleigh backscattering. This paper shows that Rayleigh backscattering (RBS) has become one of the primary noise sources in the UWFBG array-based Φ-OTDR system. We reveal the impact of the Rayleigh backscattering signal on the intensity of the reflective signal and the precision of the demodulated signal, and we suggest reducing the pulse duration to improve the demodulation accuracy. Experimental results demonstrate that using light with a 100 ns pulse duration can improve the measurement precision by three times compared with the use of a 300 ns pulse duration.
BibTeX
@inproceedings{wang2023impact,
  author = {Feng Wang and Yong Yu and Rui Hong and Ruifeng Tian and Yixin Zhang and Xuping Zhang},
  title = {{The Impact of Rayleigh Scattering in UWFBG Array-Based Φ-OTDR and Its Suppression Method}},
  booktitle = {Sensors},
  year = {2023},
  url = {https://openreview.net/forum?id=JSg6rNiVb2},
}
Figure