Compressed sensing techniques for altitude estimation in multipath conditions

In very high frequency array radars, the multipath signal and direct signal, which fall within the beamwidth of the receiving antenna, are highly correlated. This correlation degrades the performance of the low-angle direction-of-arrival (DOA) estimation in multipath conditions. By making use of the sparsity of the targets, after filtering out the clutter in the Doppler dimension, two DOA estimation approaches based on compressed sensing (CS) are proposed. The interpolated array- (IA-) CS and beamspace- (BS-) CS perform compressive sampling, respectively, on an IA and in the BS. The proposed methods are different from subspace-based methods in concept and are not subject to the restricting requirements of spatial and temporal stationarities, as well as the correlation between the sources and noise. Both simulated and measured results verify that the proposed methods provide superior performance in resolving the DOAs as compared with the spatial CS method and other conventional DOA methods.