第五章 离散傅里叶变换及其快速算法(国外英语资料).docVIP

第五章 离散傅里叶变换及其快速算法(国外英语资料) Data worth having From the usual study, accumulation and summary Where there is a problem, there must be some Please also criticize and correct me! The fifth chapter is discrete Fourier transform and its fast algorithm The derivation of 1 Discrete Fourier transform (DFT) (1) time domain sampling: Objective: to solve the problem of signal discretization Effects: the continuous signal is discretized so that the frequency spectrum of the signal is periodically extended (2) truncation in time domain: Reason: Engineering cannot handle time unlimited signals Methods: the window was intercepted by window function (usually rectangular window) Results: the time domain multiplied by the rectangular pulse signal The frequency domain is equivalent to convolution with the sampling function (3) time domain periodic extension: Purpose: to make frequency discrete The time domain is converted into periodic signals Methods: the movement in periodic extension is realized by convolution Representation: the waveform after extension can be mathematically expressed as convolution of the original waveform and the sequence of impulse sequences Results: the periodic functions after periodic extension have discrete spectrum (4) after sampling, truncation and extension The signals are discrete and periodic in both time and frequency domains The process is shown in Figure 1 Fig. 1 Schematic diagram of DFT derivation process (5) continuous time Fourier transform of processed signals: (I) discrete functions Impulse exists only at discrete frequency points Strength of The remaining points are 0 (II) is a periodic function with a period of Each cycle has a different amplitude (III) discrete time intervals (or cycles) in the time domain and reciprocal (or discrete) intervals in the frequency domain are reciprocal 2 definitions of DFT and IDFT (1) DFT definition: set is a sampling value of continuous function The width of this N point is N of DFT: (2) IDFT definition: set is