Method and system for realizing large-point FFT or DFT based on small-point algorithm kernel

The invention discloses a method for realizing large-point FFT (Fast Fourier Transform) or DFT (Discrete Fourier Transform) based on a small-point FFT or DFT algorithm kernel. Large-point FFT/DFT operation can be realized through specific peripheral operation by utilizing an existing small-point FFT/DFT algorithm kernel; according to the method, only one-dimensional FFT/DFT processing is included, matrix-free transposition operation is achieved, the structure is simple, flexibility is high, and the purchase and research and development cost of the algorithm core is reduced. 本发明公开了一种基于小点数FFT或DFT算法核实现大点数FFT或DFT的方法。利用已有小点数FFT/DFT算法核，通过特定外围运算，即可实现大点数FFT/DFT运算；本方法仅包含一维FFT/DFT处理，无矩阵的转置操作，结构简易，灵活性强，降低对算法核的采购及研发成本。