Performance analysis of frequency shift estimation techniques in Brillouin distributed fiber sensors

Area of Science:

• Optical Engineering
• Sensing Technology
• Signal Processing

Background:

• Brillouin distributed sensors are crucial for real-time strain and temperature monitoring.
• Accurate estimation of Brillouin frequency shift (BFS) is vital for sensor performance.
• Post-processing techniques significantly impact BFS estimation accuracy and efficiency.

Purpose of the Study:

• To evaluate and compare the performance of various post-processing techniques for BFS estimation.
• To identify limitations of current methods, particularly regarding truncation effects.
• To propose and validate a novel approach for improving BFS estimation accuracy.

Main Methods:

• Numerical simulations and controlled experiments were used to assess performance.
• Techniques evaluated include polynomial fitting, Lorentzian fitting, Lorentzian Cross-correlation, and Cross Reference Plot Analysis (CRPA).
• Key performance metrics include BFS uncertainty, BFS offset error, and processing time under varying SNR and truncation conditions.

Main Results:

• Lorentzian cross-correlation showed the smallest BFS uncertainty and fastest processing time.
• This technique exhibited the largest BFS offset error due to BGS truncation.
• A proposed compensation method effectively mitigated the BFS offset error.

Conclusions:

• Lorentzian cross-correlation offers high speed and precision but requires error compensation.
• The novel compensation approach significantly enhances the reliability of the Lorentzian cross-correlation technique.
• This optimized technique provides superior performance for Brillouin distributed sensing applications.