No abstract

In this study, conventional threshold way has many defects, for example, the way of extraction of signal processing for the <δν2> and experimental statistics depend on several subjective criteria during the analysis process. Therefore, we introduce another new way of using wavelet analysis that can be extracted valid signals from noise. The results of experiment indicate that it can distinguish noise from the signal <δν2>. Compared with the conventional way, the new method has better statistical distribution map of the <δν2> and smaller error.