The Design of Signal Denoising and Event Location Method based on OTDR

Optical time domain reflectometer (OTDR) is a device that uses Fresnel reflection and Rayleigh scattering principles for optical fiber detection, which is widely used in the fields of fiber characteristic measurement, fiber event localization and fiber link quality assessment. However, during the testing process, the OTDR signal is often affected by system thermal noise, scattering noise, dark current noise, and environmental noise, especially when the interference is strong, the event localization becomes difficult, which seriously affects the reliability of the detection results. In addition, the traditional event localization methods usually have the problems of complicated process and low localization accuracy. Therefore, this paper proposes an improved wavelet threshold denoising method to effectively denoise the OTDR signal. At the same time, the event localization is performed by combining the wavelet transform (WT) and cell averaging constant false alarm rate (CA-CFAR) method, so as to improve the localization accuracy and the overall performance of the system.