Alternating Projection Based Unitary Matrix Completion Method for DOA Estimation in Nonuniform Noise
Luting Lin, Xianpeng Wang, Xiang Lan, Zhiguang Han · 2022
In this paper, an effective algorithm for direction-of-arrival (DOA) estimation in the presence of uncertain nonuniform noise is proposed. The Centro-Hermitian characteristic of the covariance matrix is used to turn the complex-value matrices into real-value ones with the unitary transformation. Then a unitary matrix completion technique via alternating projection is applied to determine the noise-free covariance matrix. Finally, the DOA estimation is obtained by utilizing the unitary subspace-based algorithms. In comparison with existing algorithms, the proposed method provides better performance especially with limited snapshots. Numerical simulations demonstrate the effectiveness of the proposed method.