Deconvolution approaches applied to space-time radar data

This paper considers the operation of multi-element radar arrays in the context of Airborne Early Warning (AEW) applications. A two-dimensional convolution model is proposed to represent the transformation of data determined by the existence of targets characterized by a given relative velocity and located at a certain angle, into a corresponding Azimuth-Doppler Spectrum. The feasibility of this interpretation is demonstrated by matching of the spectrum generated through convolution with the one resulting from software simulation of the same target conditions. Two methods of discrete two-dimensional deconvolution are explored in an attempt to revert the process, obtaining an estimate of the target characteristics from simulated Azimuth-Doppler Spectra. The advantages and disadvantages of the methods are reported and the possibility of using deconvolution to preferentially retrieve target components over clutter interference is presented.