Single Snapshot Array Interpolation for Angular Estimation in Automotive Radar Applications

A lot of research efforts have been invested to im-prove angular estimation accuracy and resolution while keeping hardware efforts small. For this reason, techniques to virtually augment antenna apertures became popular. Among these are the employment of multiple-input and multiple-output (MIMO) methods or the utilization of second-order statistics in co-prime arrays. However since these virtually augmented arrays generally feature holes and thus missing spatial samples, they suffer from an increase of sidelobe levels and performance degradation. This imposes challenges to angular estimation procedures. To face those shortcomings, this paper presents a novel, fast and simple array interpolation technique to reconstruct a uniform linear array (ULA) measurement. The method is furthermore suitable for applications with severely limited snapshot count such as automotive imaging radars. Numerical experiments validate the superiority of the proposed method in terms of resolution ability, estimation accuracy and computation time compared to nuclear norm minimization (NNM) techniques.