MODELING A GTD-BASED RADAR TARGET BY THE JOINT SOFT-THRESHOLDING MAXIMUM-LIKELIHOOD METHOD

A new modeling procedure, called the joint soft-thresholding maximum-l ikelihood (JSTML) method, is proposed for radar scattering. This proce dure will cover the important practical case where the broadband scatt ering signal is received with a low signal-to-noise ratio (SNR). The J STML method uses a soft-thresholding prefiltering process to enhance S NR first, and then finds the estimates of the parameters in GTD-based model through the maximum-likelihood (ML) method. The genetic algorith m is used to overcome the issue of convergence to the local minimum. T o reduce the computational load our choice for an initial estimator is the matrix pencil method combined with the GTD-based model, which pro vides initial estimates of all parameters, including a, which parametr izes the frequency dependence of the scattering mechanism. The simulat ion results show that the JSTML method has better noise performance th an the ML method. (C) 1997 John Wiley & Sons, Inc.