EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A GAN-CNN Fusion Framework for Deep Learning-Based DOA Estimation in Low-SNR Environments

Zhenshan Zhang, Wenjie Xu iD, Haitao Zou, Shichao Yi iD

Sensors · 06.03.2026 · 26 (5) · 1676

DOI: 10.3390/s26051676 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 06.03.2026. Verfasst von Zhenshan Zhang und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science and Engineering, Jiangsu University of Science and Technology, Zhenjiang 212003, China. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Development of Key Data Algorithms for Jizhi Ship Technology.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Direction of Arrival (DOA) estimation faces significant performance degradation under low Signal-to-Noise Ratio (SNR) conditions, where traditional algorithms and deep learning models struggle due to corrupted spatial information and limited training data. To address these challenges, this paper introduces a novel two-stage framework that integrates a Generative Adversarial Network (GAN) for signal enhancement with a complex-valued Convolutional Neural Network (CNN) for DOA estimation. The proposed GAN incorporates an attention mechanism and a dedicated phase-consistent loss function to suppress noise while preserving spatial phase information critical for accurate direction finding. Enhanced signals are transformed into covariance matrices and processed by a complex-valued CNN designed to extract robust spatial features. Extensive experiments demonstrate that the proposed method achieves a DOA accuracy of 72.2% and a Root Mean Square Error (RMSE) of 3.9° at —10 dB SNR with 500 snapshots, substantially outperforming conventional and deep learning baselines. The framework also shows strong robustness to limited data, maintaining 93.8% accuracy with only 50 snapshots. The framework offers a practical solution for reliable DOA estimation in low-SNR and data-scarce environments.

Schlagwörter

Stichwörter der Autor:innen

DOA estimationarray signal processingdata augmentationdeep learninggenerative adversarial network

Autor:innen und Institutionen

  • School of Computer Science and Engineering, Jiangsu University of Science and Technology, Zhenjiang 212003, China.
  • School of Science, Jiangsu University of Science and Technology, Zhenjiang 212003, China.
  • Zhenjiang Jizhi Ship Technology Co., Ltd., Zhenjiang 212003, China.

ORCID-Profile: Wenjie Xu, Shichao Yi

Förderung

  • Development of Key Data Algorithms for Jizhi Ship Technology

Zitieren

APA

Zhang, Z., Xu, W., Zou, H., & Yi, S. (2026). A GAN-CNN Fusion Framework for Deep Learning-Based DOA Estimation in Low-SNR Environments. Sensors, 26(5), 1676. https://doi.org/10.3390/s26051676

Vancouver

Zhang Z, Xu W, Zou H, Yi S. A GAN-CNN Fusion Framework for Deep Learning-Based DOA Estimation in Low-SNR Environments. Sensors. 2026;26(5):1676. doi:10.3390/s26051676

BibTeX
@article{Zhang2026,
  author = {Zhang, Zhenshan and Xu, Wenjie and Zou, Haitao and Yi, Shichao},
  title = {A GAN-CNN Fusion Framework for Deep Learning-Based DOA Estimation in Low-SNR Environments},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {5},
  pages = {1676},
  doi = {10.3390/s26051676}
}
Kostenlos starten →