A DOA Estimation Algorithm Based on Eigenvalues Ranking Problem

Although traditional subspace direction of arrival (DOA) algorithms enable super-resolution, the algorithm performance declines sharply when the number of sources is incorrectly estimated. This article proposes a new DOA estimation method as an optimization problem similar to the conventional minimum variance distortionless response (MVDR) that fully considers the eigenvalues ranking problem. After a mathematical derivation, a concise DOA estimation expression is given based on the mapping between eigenvalues. As the algorithm considers the relative ranking of eigenvalues, the number of sources is not required for DOA estimations. Therefore, the performance of the algorithm does not degrade due to incorrect estimations in the number of sources. To enhance the algorithm performance, this article analyzes the reasons for the formation of pseudo peaks based on the eigenvalues and gives the corresponding boundary conditions. Based on the analysis of white noise, in order to overcome the influence of color noise, this article gives a way to deal with color noise. Finally, the theory of the proposed algorithm is verified through simulation experiments, and its performance is demonstrated through comparative experiments with MUSIC under white and color noise.