An adaptive fast Fourier transform demodulation algorithm for extrinsic fiber Fabry–Pérot sensor array crosstalk suppression

This paper proposes an adaptive Fast Fourier Transform (FFT) demodulation algorithm based on white light interference to address the crosstalk issue in extrinsic fiber Fabry-Perot (F-P) sensor arrays. For a single F-P sensor, the signal information is primarily distributed at the corresponding eigenfrequency. When the crosstalk occurs, the reflected spectrum resulting from the superposition of multiple sensor spectra contains multiple eigenfrequencies, along with the phase information of multiple sensors. The adaptive FFT demodulation algorithm can accurately identify all the eigenfrequencies of the reflected spectrum by adaptively adjusting the sampling length of the actual collected spectrum, and then efficiently extracts the phase-orthogonal demodulation signals corresponding to different eigenfrequencies, thereby reducing the level of crosstalk and achieving precise signal demodulation. This paper takes the two-element parallel multiplexing system of fiber optic F-P sensors as the research object, discusses the crosstalk suppression mechanism of this scheme. Through numerical simulation, the relationship between crosstalk and the initial phase of the sensing system, and length of the sensor is studied. The results show that compared with traditional FFT demodulation algorithms, the proposed adaptive FFT demodulation algorithm can reduce the fundamental frequency crosstalk (FFC) by 10-15 dB, which is below -70 dB, effectively reducing the crosstalk of the sensing array and reducing the precision control requirements of cavity length. It is expected to be applied to signal demodulation of large-scale fiber F-P sensor arrays.