Research of Signal De - noising Technique Based on Wavelet

During the process of signal testing, it is often exposed to interference and influence of all kinds of noise signal, such as data collection and transmission and so noise may be introduced. So in practical applications, before analysis of the data measured, the need for de-noising processing. The signal denoising is a method for filtering the high frequency noise of the signal and makes the signal more precise. This paper deals with the general theory of wavelet transform, the application of wavelet transform in signal de-noising as well as the analysis of the characteristics of noise-polluted signal. Matlab is used to be carried out the simulation where the different wavelet and different threshold of the same wavelet for signal de-noising are applied. An indicator of wavelet de-noising is presented, it is the indicator of smoothness. Through analysis of the experiment, considering MSE, SNR and smoothness, it can be a good way to evaluate the equality of wavelet threshold de-noising. The results show that the wavelet transform can achieve excellent results in signal de-noising; denoised signal using soft-threshold method is smoother and soft-threshold method is more suitable for more signal detail component; SNR, MSE and smoothness are all important indexes to evaluate the performance of noise reduction; threshold rule, wavelet decomposition level and wavelet function all impact the de-noising performances.