Fiber Bragg Grating Wavelength Demodulation

FBG wavelength demodulation methods include cross-correlation algorithms, cumulative sum preprocessing, spectral reconstruction using AWG, and FFT-based techniques, each offering trade-offs in accuracy, speed, and complexity.Cross-Correlation and Variable-Step Algorithms

One widely used method is the cross-correlation algorithm, often combined with a variable-step-size approach. This method interpolates the FBG spectrum to improve resolution, then calculates the cross-correlation between the measured spectrum and a reference to find the peak wavelength. A coarse search is first performed with a larger step size, followed by a fine search with a smaller step size to achieve pm-level resolution. This approach reduces computational load while maintaining high accuracy and stability, making it suitable for temperature and strain sensing in harsh environments .

Cumulative Sum Preprocessing

The cumulative sum (CUSUM) method is a simpler spectrum processing technique that reduces noise influence before determining the wavelength shift. It can be used alone or combined with other algorithms to simplify the demodulation process. While it may not significantly improve accuracy over classical methods, it offers low computational complexity and robustness against noise, making it practical for real-time applications .

Arrayed Waveguide Grating (AWG) Spectral Reconstruction

For ultra-short FBGs (US-FBGs), demodulation can be achieved using AWG-based spectral reconstruction. The AWG samples the FBG spectrum across multiple channels, and a convex optimization algorithm reconstructs the spectrum from underdetermined matrix equations. The peak of the reconstructed spectrum is then tracked to determine the wavelength. This method allows large-range demodulation with high accuracy (better than 4 pm) using only a few AWG channels, overcoming limitations of conventional AWG-based methods .

FFT-Based Cross-Correlation and Iterative Methods

Another approach uses Fast Fourier Transform (FFT)-based cross-correlation, which compares the FBG spectrum with a reference to detect shifts. Iterative improvements of this method enhance accuracy and reduce the peak-locking effect, which can occur due to limited spectral resolution. Compared to least squares fitting or centroid algorithms, FFT-based methods provide better precision and faster computation, making them suitable for dynamic sensing applications .

Summary

• Cross-correlation with variable-step size: High precision, pm-level resolution, moderate computational load.
• Cumulative sum preprocessing: Simple, noise-resistant, low computational complexity.
• AWG spectral reconstruction: Large-range demodulation, high accuracy, suitable for ultra-short FBGs.
• FFT-based cross-correlation: Fast, precise, reduces peak-locking, suitable for dynamic measurements. Each method balances accuracy, speed, and complexity, and the choice depends on the FBG type, measurement range, and application requirements.