Research On Denoising Algorithms Of Optical Time Domain Reflectometer

Optical time domain reflectometer(OTDR)as an optical fiber fault detection instrument,can measure the length of optical fiber,optical fiber transmission attenuation and fault location,etc.It has been widely used in various fields of optical fiber and cable.OTDR signal curve can reflect the loss distribution of backscattered light in optical fiber.However,due to the influence of environment and system,the event points on OTDR curve are easily submerged by noise.When the curve is seriously polluted by noise,the events in the curve are difficult to be identified.In order to solve the problem that the backscattered signal in the optical time domain reflectometer is seriously disturbed by noise,three denoising methods of OTDR signal are proposed in this thesis.The work is as follows:(1)An improved wavelet threshold denoising algorithm for OTDR signals is proposed.The traditional hard threshold wavelet denoising method in signal reconstruction will generate pseudo gibbs phenomenon,and wavelet soft threshold denoising method in reconstructing signal will bring certain reconstruction error.In order to overcome the shortcomings of the traditional wavelet soft threshold and hard threshold denoising methods,the decomposition levels,the selection of threshold and the selection of threshold function of wavelet threshold denoising are improved.The results show that the signal-to-noise ratio of the improved algorithm is significantly improved,the root mean square error is reduced,and the denoising effect is better when compared with the wavelet soft threshold and hard threshold denoising methods.(2)An OTDR signal denoising algorithm based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)and improved wavelet threshold is proposed.This method utilizes the advantages of CEEMDAN decomposition algorithm,such as anti-mode aliasing phenomenon and reduction of reconstruction error.Firstly,decompose the signal into several Intrinsic Mode Function(IMF)components.Secondly,find the critical point of IMF component dominated by noise and IMF component dominated by signal,according to the analysis method of correlation coefficient.Thirdly,remove the IMF component dominated by noise,and the IMF component dominated by signal was denoised by the improved wavelet threshold denoising method.Finally,the denoising signal is obtained from the reconstructed signal.The results show that,compared with the traditional hard threshold method,CEEMDAN-hard threshold method and the improved wavelet threshold method,the proposed method can suppress the noise better,get better de-noising effect,highlight the characteristics of OTDR events,and is easier to detect events.(3)A singular value overlapping segmentation denoising method for OTDR signals is proposed.Traditional Singular Value Decomposition(SVD)denoising in the process of signal denoising,the unified critical point selected during signal reconstruction is used as the reconstruction threshold,resulting in signal distortion or poor denoising effect of the reconstructed denoising signal.Therefore,the noisy signals are overlapped and segmented,and each segment is denoised by SVD.Singular value difference spectrum is calculated based on the singular values obtained,and each mutation point of difference spectrum is taken as the threshold of reconstruction.Compared with the traditional singular value decomposition denoising signal,the signal-to-noise ratio of the reconstructed denoising signal by the proposed is significantly improved,the useful information of OTDR signal is retained,the noise is significantly reduced,and the denoising effect is better.