Study On Fast Implementation Of Non-uniform DFT Algorithm And Its Application

In the missile guidance,it often require two-dimensional target resolution,especially in the closely radial distance,azimuths ranging from multiple flight targets for the multiple backgrounds of multiple moving targets,which can accurately identify the target.Therefore,based on the high repetition frequency broadband random frequency hopping pulse radar system,the distance dimension and velocity dimension of the echo signal can be used to identify the target in this background.In the two-dimensional processing of the echo signal,the sampled signal is regarded as the non-uniform sampling in the frequency domain,and the Fourier transform is needed to process the distance dimension and velocity dimension from the sampled data,which involves non-uniform DFT.This paper focuses on the implementation and application of this algorithm.Firstly,the background and significance of DFT,non-uniform DFT and its fast algorithm are reviewed,and its application in radar signal processing is summarized.On the basis of uniform DFT,the classical fast algorithm of non-uniform DFT and its calculation amount are analyzed,and an FPGA implementation scheme of the algorithm is given.According to the characteristics of the conventional non-uniform DFT algorithm,a fast algorithm of non-uniform DFT based on sorting interpolation is proposed.The algorithms of sorting and interpolation are analyzed theoretically and experimentally.The calculation and comparison of different combination algorithms,Combining the characteristics of Virtex6 series FPGAs and the non-uniform DFT fast algorithm,the realization scheme of the bubble sorting algorithm and the Lagrange interpolation algorithm are presented respectively,and the storage quantity and calculation quantity are analyzed.The mathematical model of high repetition rate broadband random frequency hopping pulse radar signal is given,the preprocessing process of echo signal is deduced,the two-dimensional processing of echo signal processing And other key technologies were analyzed in detail.The key technologies such as coherent reception processing,echo sampling and frame combination,distance velocity dimension processing,segment image analysis and so on in the echo signal processing are analyzed.In order to solve the problem of non-uniform Fourier transform in the two-dimensional processing,the NUFFT algorithm proposed in this paper is used to solve the imaging problem,and the effectiveness of the algorithm is verified by simulation experiments.Finally,this paper summarizes the work of the full text,and according to the completion of the paper,pointed out that the next step needs to continue to work.