An Efficient Super-Resolution DOA Estimator Based on Grid Learning

Abstract

Direction-of-arrival (DOA) estimation based on sparse signal reconstruction (SSR) is always vulnerable to off-grid error. To address this issue, an efficient super-resolution DOA estimation algorithm is proposed in this work. Utilizing the Taylor series expansion, the sparse dictionary matrix is constructed under the off-grid model. Then, a polynomial optimization function is established based on the orthogonality principle. By minimizing the given objective function, we derive an efficient closed-form solution of the off-grid errors. Using the estimated off-grid errors, the discretized grid can be iteratively learned and approaches the true DOAs. With the newly learned grid, accurate DOA estimations can be achieved through the SSR scheme. The proposed algorithm converges fast and achieves precise DOA estimations even the step size of the discretized grid is large. The superior performance of the proposed algorithm is demonstrated by the simulation results.