Sparse Bayesian Covariance Refinement for Direction Finding with 1.5-Bit Cross-Dipole Arrays

Sparse cross-dipole arrays with 1.5-bit analog-to-digital converters provide an economical way to estimate the directions of arrival (DOAs) of polarized electromagnetic sources. The existing estimator reconstructs the covariance matrix of the array from the three-level measurements and then estimates the DOAs by parallel-co-array MUSIC (PC-MUSIC). However, the reconstructed covariance matrix contains finite-snapshot errors, which limit the DOA estimation performance. In this paper, we propose a polarimetric covariance sparse Bayesian inference (PC-SBI) method to refine the reconstructed covariance matrix before DOA estimation. The four reconstructed covariance blocks are modeled jointly as a sum of the contributions from a set of candidate directions, each with a two-by-two covariance block describing the powers and the correlation of the two dipole components. Based on a Gaussian covariance pseudo-likelihood, a generalized expectation–maximization algorithm is derived to update the blocks, the directions and the noise variance, in which the directions are updated using the exact steering vector without any angular grid. Our numerical results show that PC-SBI improves the DOA estimation performance of PC-MUSIC significantly at moderate and high signal-to-noise ratios (SNRs), but its computational cost is considerably higher. With the default parameters of the experiments and an SNR of 10 dB or higher, it even outperforms PC-MUSIC applied to the unquantized measurements, whose error is well above the Cramér–Rao bound. This gain comes from the covariance fit, since quantization cannot add information. For well-separated sources, a similar accuracy, in some cases a better one, is obtained by applying the existing sparse and parametric approach to the sum of the two auto-covariance blocks, and the joint fit of the four blocks is clearly more accurate only for closely spaced polarized sources.

Authors

Institutions

Publication Details

Journal
Eng—Advances in Engineering
Published
2026-11-18
DOI
https://doi.org/10.3390/eng7110657
Primary Topic
Direction-of-Arrival Estimation Techniques
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Sparse Bayesian Covariance Refinement for Direction Finding with 1.5-Bit Cross-Dipole Arrays

Zhenyu Liu, Zheng Lu, Di Song, Zhiyong Cheng et al.
Eng—Advances in Engineering
Direction-of-Arrival Estimation Techniques
article

Sparse Bayesian Covariance Refinement for Direction Finding with 1.5-Bit Cross-Dipole Arrays

Zhenyu Liu, Zheng Lu, Di Song, Zhiyong Cheng, Yu Wang
article en

Abstract

Sparse cross-dipole arrays with 1.5-bit analog-to-digital converters provide an economical way to estimate the directions of arrival (DOAs) of polarized electromagnetic sources. The existing estimator reconstructs the covariance matrix of the array from the three-level measurements and then estimates the DOAs by parallel-co-array MUSIC (PC-MUSIC). However, the reconstructed covariance matrix contains finite-snapshot errors, which limit the DOA estimation performance. In this paper, we propose a polarimetric covariance sparse Bayesian inference (PC-SBI) method to refine the reconstructed covariance matrix before DOA estimation. The four reconstructed covariance blocks are modeled jointly as a sum of the contributions from a set of candidate directions, each with a two-by-two covariance block describing the powers and the correlation of the two dipole components. Based on a Gaussian covariance pseudo-likelihood, a generalized expectation–maximization algorithm is derived to update the blocks, the directions and the noise variance, in which the directions are updated using the exact steering vector without any angular grid. Our numerical results show that PC-SBI improves the DOA estimation performance of PC-MUSIC significantly at moderate and high signal-to-noise ratios (SNRs), but its computational cost is considerably higher. With the default parameters of the experiments and an SNR of 10 dB or higher, it even outperforms PC-MUSIC applied to the unquantized measurements, whose error is well above the Cramér–Rao bound. This gain comes from the covariance fit, since quantization cannot add information. For well-separated sources, a similar accuracy, in some cases a better one, is obtained by applying the existing sparse and parametric approach to the sum of the two auto-covariance blocks, and the joint fit of the four blocks is clearly more accurate only for closely spaced polarized sources.

Eng—Advances in EngineeringVol. 7(11)
Nanjing University of Science and Technology (CN), Nanyang Normal University (CN), Chaohu University (CN)
Anhui Provincial Department of Education, Chaohu University
Direction-of-Arrival Estimation Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon