Research And Module Realization Of Time-frequency Analysis Method Based On Wavelet Transform

The modern electronic information technology is developing towards a high-speed and wide-bandwidth direction,it continually improves its requirement for systematic index,meanwhile,it also puts forward a new request for comprehensive analysis of signals.Only from time domain or frequency domain to analyze cannot meet peoples' extraction of signal characteristics.This study will start from a mixed domain instrument to slove the limitation that the traditional oscilloscopes just collect,resume and analyze signals from a single signal domain,and it can simultaneously analyze the time domain,frequency domain and time-frequency joint domain to form a multidimensional measuring instrument to measure signals.Therefore,multi-domain analysis is an indispensable part of current signal characteristic analysis.Under this requirement,from the perspective of multi-information analysis,this study aims to construct a multi-domain analysis instrument from the time domain,frequency domain,modulation domain and time-frequency joint analysis domain.This study's topic mainly from the time and frequency domain,regarding the time-frequency analysis through the wavelet transform as its main line,will state the theme of the wavelet transform from the aspects of theory and implementation,at the same time,it also will combine the digital down-conversion with the wavelet transform to analyze signals more flexibly.The concrete contents are following:1.Stating the theory of the wavelet transform and reproducing the implementation structure commonly used by the wavelet transform,evaluating its model of implementation,including the comparisons among the resource consumption,calculation speed and analysis accuracy and so on.2.Improving the implementation model of the wavelet transform.The main limited factors of implementing the wavelet transform is its large resource consumption and long calculating time.Mixing and utilizing the commonly used methods of the wavelet transform,optimizing its resource consumption of realization,and comparing its effects of realization structure with that of commonly used.3.Implementating the improved scheme of the wavelet transform on the FPGA.Analyzing the quantization results of partial parameters and contrasting the the accuracy of the wavelet transform realized on the FPGA.4.Combining the digital down-conversion and the wavelet analysis can change the relationship between the signal frequency and the sampling frequency through the digital down-conversion;the time-domain and the frequency-domain resolution of different analysis frequency points can be changed as well,and make the the analysis of the wavelet transform more flexible.The realization of the digital down-conversion utilizes the CORDIC algorithm to shift the frequency spectrum of signals;uses the CIC filter to realize the high-efficiency,multiplicity-ration decimation filter;takes advantage of the coefficient characteristics of half band filter and the transition characteristics to achieve double decimation filter,finally the function of digital down-conversion will be achieved.The wavelet transform of digital down-conversion can change the time-frequency resolution of analyzing frequency points and increase the time of the wavelet transformat analysis.Through investigating the above questions,Realize the time-frequency joint analysis function based on the wavelet transform,in the meanwhile,through the combination of the digital down-conversion and the wavelet transform to achieve more flexible and efficient time-frequency analysis of signals.