Unambiguous Peak Recognition for a Silicon Fabry-Pérot Interferometric Temperature Sensor

Wavelength tracking method is widely used for fiber-optic sensors based on Fabry-Pérot interferometers (FPIs). Although wavelength tracking provides high resolution, it encounters the well-known problem of direction ambiguity when the spectrum shifts by more than half of the free-spectral range. In this paper, we propose a new signal processing method that can unambiguously recognize the spectral peaks of a silicon FPI by obtaining their unique order values to overcome the direction ambiguity and achieve measurement over extended range. New elements of the proposed method include: 1) an average peak order calculated from multiple peak pairs is used to reduce variation of the calculated order number, and 2) a modified dispersion model of silicon is involved in the calculation of the peak order to suppress drift of the calculated peak order. The averaging method generally decreases the data processing rate, but does not affect the response time of the sensor. We experimentally demonstrate the feasibility of the proposed method using our recently developed fiber-optic temperature sensor using a 200-μm-thick silicon FPI. Results show that the measurement of temperature over 0 to 100 °C is realized, leading to an increase of more than five times in dynamic range.