Comparison of Sparse Coprime-Related Array Configurations for Automotive Radar Applications

Abstract

Two sparse multiple-input multiple-output (MIMO) radar configurations are compared against a conventional dense MIMO radar to investigate low-complexity, super-resolution direction-of-arrival (DOA) estimation from a single snapshot. Both sparse configurations divide the virtual array (VA) into two uniformly spaced subarrays with a coprime relationship between them. The first is the well-known coprime array in which the inter-element spacings of the two subarrays are coprimes. The second is the VEXPA configuration, where both subarrays share the same inter-element spacing and a coprime parameter defines the relative shift between them. The analysis highlights the spatial resolution gains of sparse configurations over dense arrays while revealing limitations from the de-aliasing mechanism and reduced degrees of freedom (DOFs) inherent to the subarray partitioning.