Low-Rank STRAP Filter Via Alternative Unfolding HOSVD for FDA-MIMO Radar

Abstract

The multidimensional structure of frequency diverse array (FDA) multiple-input–multiple-out (MIMO) radar signals has attracted a lot of attention. It allows to extend conventional space–time adaptive processing to space–time-range adaptive processing (STRAP). In this letter, we propose two tensorial filters for FDA-MIMO-STRAP, called the clutter subspace filter and the clutter-free subspace filter, which exploit the low-rankness of the clutter to achieve better clutter suppression in a small auxiliary training data scenario. The proposed method makes use of the alternative unfolding higher order singular value decomposition with different dimensional partitions. Numerical results demonstrate the effectiveness of the proposed filters over existing low-rank vectorial and tensorial methods.