Comparative Study on Noise Reduction Effect of Fiber Optic Hydrophone Based on LMS and NLMS Algorithm
Zhihua Yu1,2, Yunfei Cai1,2, Daili Mo1,2
1School of Automation, China University of Geosciences, Wuhan 430074, China.
Sensors (Basel, Switzerland)
|January 18, 2020
Summary
This study compares adaptive filtering techniques for fiber-optic hydrophones. The least mean square (LMS) algorithm demonstrated superior performance over the normalized least mean square (NLMS) algorithm in denoising signals.
Area of Science:
- Signal Processing
- Optical Fiber Sensing
- Acoustic Sensing
Background:
- Adaptive filtering offers real-time processing, low computational complexity, and robustness, making it valuable in diverse fields like communication and sensing.
- Interferometric fiber-optic hydrophones are crucial for acoustic sensing but susceptible to noise.
- Denoising is essential for accurate signal interpretation in fiber-optic sensing applications.
Purpose of the Study:
- To evaluate and compare the effectiveness of two adaptive filtering algorithms, Least Mean Square (LMS) and Normalized Least Mean Square (NLMS), for denoising interferometric fiber-optic hydrophone signals.
- To assess the performance of LMS and NLMS algorithms in terms of reducing harmonic distortion and improving signal quality.
Main Methods:
- Implemented a reference channel using an interferometer with identical parameters to the signal channel.
- Applied two adaptive filtering schemes: the Least Mean Square (LMS) algorithm and the Normalized Least Mean Square (NLMS) algorithm to the hydrophone's sensing signal.
- Utilized comparative analysis to evaluate denoising performance based on key metrics.
Main Results:
- The Least Mean Square (LMS) algorithm exhibited superior performance compared to the Normalized Least Mean Square (NLMS) algorithm.
- LMS demonstrated greater effectiveness in reducing total harmonic distortion.
- Significant improvements in signal-to-noise ratio and overall filtering effect were observed with the LMS algorithm.
Conclusions:
- The LMS algorithm is a more effective choice for denoising interferometric fiber-optic hydrophone signals than the NLMS algorithm.
- Adaptive filtering, particularly using the LMS algorithm, enhances the reliability and accuracy of fiber-optic hydrophone sensing.
- The proposed method of using a reference channel alongside adaptive filtering significantly improves signal quality in optical fiber sensing.
