Compressed sensing radar imaging of off-grid sparse targets

Compressed sensing (CS) technique has been applied to radar imaging to maintain the imaging resolution with reduced amount of data. A necessary step for computation in existing CS radar imaging regimes is to divide the scene of interest into discrete grids, and the scattered signal from each grid is approximated as being reflected from an ideal point scatterer at the center of that grid. However, this approximation significantly affects CS radar imaging performance due to scatterers not in the centers, or off-grid. Existing algorithms utilize atom optimization or iterative optimization to image those off-grid scatterers. However, these algorithms require a minimal distance between scatterers. In this work, by redesigning the imaging scheme, modifying transmitted waveform and processing procedure, with the same amount of data, suitable bases can be selected and the imaging resolution is refined. The off-grid targets are then imaged on finer grids. Experiments show the performance and robustness of the scheme under different SNRs.