Non-Iterative DOA Estimation in Time-Modulated Array under Unidirectional Phase Center Motion

Existing direction of arrival (DOA) estimation methods based on the unidirectional phase center motion (UPCM) in time modulated linear array (TMLA) mainly face two issues: the redistribution of noise and the limited range of angle observation. In this letter, the orthogonality between the array manifold and the noise subspace is analyzed with respect to the time schemes and the number of harmonics, leading to an optimal UPCM scheme with non-uniform noise. Furthermore, a noniterative two-stage method is proposed to address the performance degradation under such a scheme. In the first stage, the rough estimates of noise subspace are obtained by the eigen value decomposition (EVD) of the array covariance matrix (ACM) directly. Next stage, the updated estimates of noise subspace are further obtained by performing the generalized EVD on the ACM and noise covariance matrix. Numerical simulations demonstrate that the proposed non-iterative subspace-based method can provide excellent estimation accuracy with reduced complexity. Meanwhile, the proposed method also provides similar estimation performance even with gain-phase errors and location errors.