High-Order Polynomial Fitting Assistance for Fast Double-Peak Finding in Brillouin-Distributed Sensing

Area of Science:

• Optical Fiber Sensing
• Photonics
• Signal Processing

Background:

• Accurate retrieval of Brillouin frequency shifts (BFS) is crucial for distributed optical fiber sensing.
• Double-Gaussian fitting is commonly used but computationally intensive, especially for spectra with dual peaks.
• Existing methods face challenges in speed and accuracy under varying signal-to-noise ratios (SNRs) and spectral complexities.

Discussion:

• A novel high-order polynomial fitting method is introduced to accelerate double-Gaussian fitting for dual-peak BFS retrieval.
• This method provides a reliable initial estimation for dual local BFS values using a sixth-order polynomial.
• The polynomial estimation serves as initial parameters for nonlinear double-Gaussian fitting, enhancing convergence.

Key Insights:

• The proposed polynomial fitting method reduces iterations by 4.9-fold.
• Processing time is improved by 3.4-fold compared to standard double-Gaussian fitting.
• Experimental validation confirms reliability across different SNRs and spectral scenarios in distributed Brillouin sensors.

Outlook:

• Potential for real-time BFS monitoring in advanced fiber optic sensing applications.
• Further optimization of polynomial order and fitting algorithms could yield greater computational efficiencies.
• Applicability to other spectral analysis techniques requiring initial parameter estimation.