The Wavelet Spectrum Method In The Analysis Of 1/f Noise In Electronic Device

The noise inside the electronic devices (especially the low-frequency noise such as 1/f noise) is the key factor limiting the sensitivity and precision of the devices. And at the same time, it is the important sensitive parameter to characterize the quality and reliability of the devices. Therefore, noise has been studied and used widely as the diagnostic and evaluating tool.Up till now, the power spectral density is the traditional factor applied to characterizing the low-frequency noise in both the experiments and the theories. Therefore, theory and experiment has proved power spectrum based on FFT is not an entire sign for noise recently. Wavelet analysis combine the frequence domain and time domain analysis has advantage to entire sign noise. Noise procession with wavelet must be a main stream method. Although the wavelet spectrum analysis is the way newly used, in order to analyse different types of noise and obtain satisfied results, the research for different kinds of noise with the frequency-domain analytical method has become the subject in front of us.This paper is focused on the analysis to different types of noise from the wavelet spectrum evaluation and wavelet spectrum fitting. The best algorithms are given, and the feasibility of the algorithms are proved. Then compared with the Fourier spectrum, the compare not only between wavelet spectra and Fourier spectra, but also the spectra estimation. Wavelet spectra plots a linear spectrum and the variance of it is little . we prove that the wavelet spectra is superior to the traditional Fourier spectra. The application shows that these softwares for the frequency-domain analysis of the electronic devices can abtain a satisfied result. It is considered that thses softwares will have a good practical utility.