A Motion Parameter Estimation Method for Radar Maneuvering Target in Gaussian Clutter

It is known that accurate motion parameter estimation makes significant benefits to radar target tracking, imaging and recognition. However, maneuvering targets usually cause across-range-unit and across-Doppler-unit effects, which make it difficult to accurately estimate the motion parameters of the target. The generalized Radon-Fourier transform (GRFT) has been proposed to estimate the motion parameter vector of maneuvering targets, but only the noise scenario is considered. In this paper, the normalized adaptive GRFT (NAGRFT) is first proposed for motion parameter estimation, which is proven to be the maximum likelihood estimator (MLE) of the motion parameter vector in Gaussian clutter. Then, we derive the Cramer-Rao lower bound (CRLB) for the motion parameter vector as a comparison of the estimation mean square error (MSE) of the NAGRFT. In this derivation, the target motion is described by a high-order polynomial model, and a Gaussian-shaped function is adopted to characterize the power spectrum density of the homogenous clutter. Finally, to demonstrate the optimal estimation performance of the proposed NAGRFT, numerical experiments are provided, which show that the estimation MSE of the NAGRFT can reach the CRLB. Experiments are also provided to analyze the relationship between the CRLB and clutter parameters.