Study On Velocity Measurement Algorithm Of Wideband Radar Based On Adjacent Cross-correlation

In modern radar systems,wideband radars have a lot of advantages such as high range resolution,strong anti-clutter capability,and high parameter estimation accuracy.Therefore,the moving target parameter estimation algorithm has great practical significance.At present,Linear Frequency Modulated signal has been generally used by wideband radars.And de-chirping processing was usually used by wideband radars to perform pulse compression.In this paper,we develop a parameter estimation algorithm for moving targets and conduct an in-depth study in a velocity measurement scenario.The main content of this paper includes:1.The detailed research of the principle of de-chirping processing,wideband velocity estimation,wideband angle estimation and noise reduction in the actual wideband radar signal processing flow is expanded.2.The expression of Fourier transform is derived for wideband signal model,then Rife algorithm and phase estimation algorithm are presented.After that,we improve the J-T algorithm,and verify its performance through simulation.Aiming at the zero-filling of signals in practical applications,a frequency estimation algorithm based on fast bisection is proposed.Results shows that the estimation accuracy is close to the Cramer-Rao bound with a small amount of calculation.Then,the performance is compared with the existing methods through simulation experiments.3.Based on the wideband velocity estimation scenario,the traditional wideband parameter estimation algorithm is introduced.Then,this paper proposes a frequency estimation and phase ambiguity reduction algorithm based on envelope cross-correlation.After that,this paper compares it with a traditional method.The results shows that the new method that proposed by this article makes some improvements in variance of estimated results.Finally,a general-purpose de-noising reconstruction algorithm is given,and its performance is compared with the algorithm without de-noising in simulation experiments.From the results,we can know that it can improve the estimated varience at the cost of longer runtimes.