Study On Motion Parameter Estimation Of Intensive Multi-targets And Radar Imaging Based On Capon Method

Resolution is one of the important metrics to measure the effectiveness of such radar applications as motion parameter estimations of intensive multi-targets and radar imaging.Comparing with traditional digital processing methods,super resolution signal processing techniques can provide beyond-limitation resolution and robust ability of unwanted interference or side lobes suppression.In our paper,faced with the problems still existed in the modern radar,mainly including 1)motion parameter estimation of intensive multi-targets,whose estimation precisions are affected by the superposition of interference or side lobes,2)target identification of the multi-scattering point target,as well as 3)the finite samples caused by the no-recurring radar scenes,we first transform the problems of motion parameter estimations of intensive multi-targets and radar imaging into spectrum estimation problems.And then,on the premise of finite samples,the study on super resolution methods are carried out based on the theory of Capon filtering banks.The researches mainly contain the following contents:1.Briefly introduce the radar wideband and narrow-band signal models,respectively.The problems of motion parameter estimations of intensive multi-targets and radar imaging are transformed into spectrum estimation problems.2.Given the classic spectrum estimation problem of two-dimensional complex exponential signal,the matched filtering based maximum likelihood estimation(MLE)method,and the beamforming based Capon filtering banks method are introduced.Besides,the Cramer-Rao bound is derived,and the performances are compared between the MLE and the Capon methods,in the aspects of parameter estimation accuracy and the first side lobe attenuation.3.On the premise of finite samples,in order to obtain the high-precision and high-resolution parameter estimation and imaging results,we study the rank-reducing theory from the high definition vector imaging(HDVI)method and propose the Eigen structure and subspace based Capon method(ES-Capon).For the single-frequency and multi-frequency component exponential signals,the experiments are respectively performed to compare the performance of the ES-Capon and MLE method.And the used measurement metrics include parameter estimation accuracy,the first side lobe attenuation and resolution.The simulation results show that the ES-Capon method has better resolution and estimation accuracy.4.For the computation complexity problems with the ES-Capon method,the 2-dimensional FFT(2-d FFT)based fast implementation method,and the 2-dimensional Chirp-Z transform(2-d CZT)based fast implementation method are proposed.5.We introduce the generation techniques of the echo waveforms for the narrow-band and wideband radar,respectively.Using MF and Dechirp method,the problems of motion parameter estimations of intensive multi-targets and radar imaging are simplified as spectrum estimation problems.Then the ES-Capon method is applied in these two mentioned applications.The simulations and real data experiments are performed to verify the effectiveness of the ES-Capon method.