Resolution limits and efficient signal processing for fiber optic Bragg grating sensors with direct spectroscopic detection

摘要

Theoretical resolution limits of evaluation of centroid shift in fiber optic Bragg grating sensors are analyzed. In the theoretical model, it is assumed that random additive Gaussian noise in a sensor signal acquisition equipment is the main origin of the system error. To provide noise resistant demodulation, a new optimal signal processing algorithm is suggested. It allows accurate detection of the Bragg wavelength shifts within large range and keeps resolution very close to theoretical limit imposed by a random noise. The new demodulation technique is also compared with well known centroid detection algorithm.