Randomized step frequency radar with adaptive compressed sensing

We develop a novel algorithm for range and velocity joint estimating in randomized stepped frequency radar. By exploiting sparseness of the targets and invoking compressed sensing (CS) theory, higher resolution can be achieved. However, in actual radar system, some previous CS methods probably suffer from off-grid point effect, which causes performance degradation of the methods. We produce an adaptive CS method to mitigate the off-grid point effect. The performance of the proposed algorithm on accuracy and resolution of range-velocity joint estimate are illustrated by simulations and field experiments.