Shift Happens, So Filtering for DTS Matters

This paper tells about the problem of choosing a signal processing method for a fiber-optic temperature sensor based on Raman scattering. The fiber optic sensor is inherently connected to the optoelectronic transformation, which leads to a low signal-to-noise ratio. A lot of well-known filtering algorithms can result in deterioration in the characteristics of the device due to this one. It is especially noticeable when there are sharp changes in the signal amplitude. The paper presents the results of signal processing from a real Raman sensor. A laboratory bench has been designed that allows heating the several fiber areas. Two cases are being considered. At first, the temperature profile contained two areas, all of which consisted of three sharp temperature drops. It allowed us to demonstrate how the slightest deterioration in spatial resolution after signal processing negatively affects the device characteristics. The second situation is the heating of a fiber region significantly exceeding the spatial resolution of the device that is less difficult to process.