Fast Direction-of-Arrival Estimation via Coarray Interpolation Based on Truncated Nuclear Norm Regularization

Abstract

Non-uniform linear arrays have the ability to provide higher degrees of freedom (DOF) than conventional uniform linear arrays (ULAs) by making use of their virtual coarray. However, some non-uniform arrays such as coprime arrays have holes in their virtual array. In order to utilize the full DOF of such arrays, coarray interpolation is performed. Current interpolation techniques are mostly based on nuclear norm minimization of the coarray covariance matrix. However, these interpolation techniques are computationally inefficient, especially for larger matrices. In this brief, we propose a coarray interpolation method based on truncated nuclear norm minimization. This method approximates the rank of the matrix better and is computationally much more efficient as it uses a weighted iterative approach to interpolate a matrix. In numerical simulation, we perform direction of arrival (DOA) estimation using the proposed method and demonstrate its effectiveness both in terms of computation time and accuracy.