Crosstalk in the signal processing of a cascaded Fabry-Perot and a fiber Bragg grating sensor

Current demodulation methods for fiber Bragg grating (FBG) peaks and Fabry-Perot (FP) extremums rely either on mathematical approaches or on physical models tailored to a single sensing mechanism. As a result, they inherently fail to distinguish reciprocal errors arising from spectral crosstalk between FBG and FP signals in cascaded FBG-FP sensor systems. The asymmetric influence of FP fringe slopes on the FBG peak position, as well as FBG side-lobes effect on FP peak or dip distortion, can introduce errors in identifying the wavelength positions of FBG peaks and FP fringes. In this work, we developed a model to analyze spectral errors in cascaded FBG-FP sensors and validated it through both analytical and experimental approaches. For the analytical method, the errors in FBG peak were calculated within one free spectral range (FSR) of the FP fringe shift for two representative FP configurations, considering influences of the FBG bandwidth and its peak reflectivity. The errors in FP peaks, on the other hand, were extracted as a function of their relative spectral position to the main FBG lobe. For experimental validation, a 10 mm apodized FBG cascaded with a 100 µm thick silicon pillar at the fiber tip was employed to investigate the spectral errors in FBG and FP peaks as predicted by the model. The proposed approach offers a practical framework for characterizing spectral distortions in such systems, enabling accurate parameter measurements and facilitating high-resolution sensing performance.