High-resolution range-doppler imaging by coherent block-sparse estimation

We consider a pulsed radar that illuminates a sparse scene of point-like scatterers. Formulating the data model for the receive signal in Fourier space leads to a block sparse estimation problem for the range profile spectra of each Doppler channel. Then, in each Doppler channel, we compute the delays using a parametric high-resolution method. The mixed sparse/parametric approach overcomes the disadvantages of matched filtering concerning resolution, artifacts, and Doppler ambiguities and has less computational complexity than full two-dimensional sparse estimation methods. We show results from numerical simulations and experimental measurements for passive bistatic radar using GSM as illuminator of opportunity.