一种改进的小波去噪方法在OTDR中的应用

Application of an improved wavelet denoising method in OTDR

光时域反射仪(OTDR)测试中,系统链路噪声严重影响测试准确度,尤其是对于0.05~0.2 dB熔接损耗的准确检测,需要提高测试曲线的信噪比。根据OTDR波形特点,在小波阈值去噪的基础上提出一种改进的小波去噪方法,改进阈值利用对数函数的非线性并引入噪声方差,更利于滤除噪声;改进阈值判断函数留下原始数据细节并滤除噪声,且不改变反射事件和非反射事件的波形属性。实际测试表明:测试曲线相对信噪比提升了10 dB以上,去噪后配以曲线平滑滤波和恒虚警(CFAR)检测,准确检测出熔接事件。

In optical time domain reflectometer(OTDR)test,the system link noise seriously affects the test accuracy,especially for the accurate detection of 0.05-0.2 dB fusion loss,it is necessary to improve the signal-to-noise ratio of the test curve.According to the characteristics of OTDR waveform,based on wavelet threshold denoising,an improved wavelet denoising method is proposed.The improved threshold makes use of the nonlinearity of logarithmic function and introduces noise variance,which is more conducive to filtering out noise;the improved threshold judgment function leaves original data details and filters out noise,and does not change the waveform attributes of reflection events and non-reflection events.The actual test results show that the relative signal-to-noise ratio of the test curve is increased by more than 10 dB,and the fusion event can be accurately detected after denoising with curve smoothing filter and constant false alarm rate(CFAR)detection.