Frequency Shifts Estimation for Sensors Based on Optoelectronic Oscillators

Abstract

Optoelectronic oscillators (OEO) based sensors generate low-phase-noise signals, whose frequency is in the order of GHz. These sensors are highly sensitive to parameters like strain, transverse load or temperature. Particularly, an OEO transforms a wavelength change into a frequency shift in the order of MHz. This creates a particular challenge for measuring instruments used in OEO, where novel mathematical methods for advanced processing of their signals is required. The principle of rational approximations is a frequency measurement technique based on number theory, with the fundamental property of measurement time dependent on the frequency value of input signals. This means that if the frequency of input signals increases, less time for measuring is required. This is quite useful for time-frequency measurement systems, where if more accuracy is needed, more time is required for measuring. This paper shows how the principle of rational approximations can be used in OEO based sensors, this allows to estimate the frequency shifts in a bandwidth from 6 to 600 MHz, with a maximum uncertainty of 2.5 MHz. Such an accuracy is achieved in a time as short as 0.2 μs.