Research On Multi-Target Detection And Parameter Estimation Based On Large Gazing Field Of View

In order to adapt to complicated environment,diverse targets and multiple missions,modern radar systems have great challenges in multi-target processing.In our dissertation,some complicated problems still existed in multi-target processing,mainly including 1)inter-targets' shadowing effect which exists in the detection of intensive multi-target,range migration of high-speed targets,2)great computational complexity of multi-target joint range/velocity estimation,as well as 3)poor performance on angle estimation of classical method in unresolved targets scenario.Consider a gazing radar system that transmits wide beam to obtain large field of view.The study on multi-target detection and parameter estimation technology are carried out based on the gazing radar system.In this dissertation,the researches mainly contain multi-target detection,range/velocity estimation and angle estimation.(1)Briefly introduced the radar signal model and the principle of pulse compression based on matched filtering.According to the principle of amplitude comparison monopulse,the hybrid outputs of sum and difference channel signals are corrected by beam pattern gain.(2)In order to solve the range migration of high-speed targets,a pre-processing method based on coast/fine velocity compensation is presented.Besides,a modified CFAR(M-CFAR)detection method is introduced,which obtains more accurate estimation of interference power by rejecting the extremum in reference windows,and makes the edges of signal also satisfy operation conditions of CFAR detection by extending the two-dimensional signal.(3)For the computation complexity problems of maximum likelihood estimation(MLE)method,a multi-target inter-partition estimation based on MLE method and a CZT-IFFT based fast implementation method are proposed.Besides,a MLE based method is introduced to solve the problem of velocity ambiguity,whose feasibility is verified by simulation experiments.(4)When more than one target are closely spaced in the range within the radar beam,the traditional amplitude comparison monopulse(ACM)method is incapable of resolving them and fail to work.Therefore,a multi-target amplitude comparison based on maximum likelihood(MAC-ML)method is introduced,which uses CLEAN algorithm to weaken the interference of strong target component in channel signals.Besides,the Cramer-Rao lower bound(CRLB)is derived and the simulation results show that the performance of MAC-ML method can achieve CRLB.