Fast and Accurate Single-Snapshot DOA Estimation: Iterative Interpolated Approaches

Estimating the Direction of Arrival (DOA) of a single source signal from a single observation of an array data still plays an important part in practical application scenarios. To address the problem, this letter proposes two iterative, fast and accurate approaches namely Q-Shift based DOA Estimatio...

Full description

Saved in:
Bibliographic Details
Published in:IEEE geoscience and remote sensing letters 2023-01, Vol.20, p.1-1
Main Authors: Liu, Yicen, Zhao, Peiyan, Yao, Denghui, Liu, Bo, Zhang, Junning
Format: Article
Language:English
Subjects:
Algorithms
Asymptotic properties
Coefficients
Cramer-Rao bounds
Cramér-Rao bound
DFT interpolation
Direction of arrival
direction of arrival (DOA) estimation
Direction-of-arrival estimation
Discrete Fourier transforms
Error functions
Estimation
Estimation accuracy
Fourier transforms
Interpolation
Iterative algorithms
Iterative methods
Radar signal processing
Signal processing algorithms
Superresolution
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Estimating the Direction of Arrival (DOA) of a single source signal from a single observation of an array data still plays an important part in practical application scenarios. To address the problem, this letter proposes two iterative, fast and accurate approaches namely Q-Shift based DOA Estimation Algorithm (QS-DOAEA) and Tradeoff between A&M and Q-Shift based DOA Estimation Algorithm (TAQ-DOAEA). QS-DOAEA adopts the interpolation of shifted DFT coefficients to iteratively obtain the near-optimal estimation. TAQ-DOAEA employs the error function by simultaneously mixing the q -shifted and half-shifted DFT coefficients, which only requires two iterations. Numerical results reveal that QS-DOAEA achieves significant improvement in terms of estimation accuracy and TAQ-DOAEA expedites the convergence. The proposed estimation algorithms demonstrate that they have an asymptotic variance that is at least 1.0013 times asymptotic Cramér-Rao bound (ACRB), and provide more than 80× reduction in the computational cost, compared to the baseline method. All our proposed algorithms and code are available at https://github.com/jn-z/.
ISSN:1545-598X
1558-0571
DOI:10.1109/LGRS.2023.3283288