Deep Learning-Based Multipath DoAs Estimation Method for mmWave Massive MIMO Systems in Low SNR

Yu, Jiarun; Wang, Yafeng

doi:10.1109/tvt.2023.3239402

Cited by 16 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(e) the simple CNN network proposed in [28], called CNNsimple. (f) the MIMO CNN network proposed in [30], called CNNMIMO.…”

Section: The Algorithm Used For the Comparisonmentioning

confidence: 99%

“…In [29], multilayer CNN is used for source number and DOA estimation, which has a good DOA estimation effect, and there is no need to adjust SNR and the number of snapshots, which makes the on-grid model more universal. Literature [30] describes a method for estimating the DOA in millimetre-wave multiple-input multipleoutput (MIMO) systems using CNN without prior knowledge of the number of multipaths. Literature [31] introduces 2D DOA estimation algorithms for dilated convolutional networks that can adapt to low elevation angles.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A DOA Estimation Method Based on an Improved Transformer Model for Uniform Linear Arrays with Low SNR

Wang,

Zhou,

et al. 2024

IET Signal Processing

View full text Add to dashboard Cite

In this paper, the Star-Transformer model is improved to obtain more accurate direction of arrivals (DOA) estimation of underwater sonar uniform linear array (ULA) under low signal-to-noise ratio (SNR) conditions. The ideal real covariance matrix is divided into three channels: real part channel, imaginary part channel, and phase channel to obtain more input features. In training, the real covariance matrix is used under different SNRs. In testing, the covariance matrix of samples in the real environment is used as input. The on-grid form is used to estimate the DOA of multiple signal sources, which is modelled as a multilabel classification problem. The results show that the model can be effective and can still have a good DOA estimation performance under the conditions of trained and untrained SNRs, different snapshots, signal power mismatch, different separation angles, signal correlation, and so on. It shows that the model has excellent robustness.

show abstract

“…(e) the simple CNN network proposed in [28], called CNNsimple. (f) the MIMO CNN network proposed in [30], called CNNMIMO.…”

Section: The Algorithm Used For the Comparisonmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A DOA Estimation Method Based on an Improved Transformer Model for Uniform Linear Arrays with Low SNR

Wang,

Zhou,

et al. 2024

IET Signal Processing

View full text Add to dashboard Cite

show abstract

“…Lota et al [43] present an exploration of 5G uniform linear arrays, employing beamforming and spatial multiplexing for outdoor urban communication at various GHz frequencies, showcasing a robust two-level approach. Researchers in [44], [45] leverage deep networks for DoA estimation in low SNR scenarios, demonstrating the adaptability of DL techniques in antenna arrays, whereas in [46] there is a focus on near-field DoA estimation in MIMO systems by a complex ResNet framework.…”

Section: Prior Workmentioning

confidence: 99%

A Novel Three-Dimensional Direction-of-Arrival Estimation Approach Using a Deep Convolutional Neural Network

Mylonakis,

Zaharis

2024

IEEE Open J. Veh. Technol.

View full text Add to dashboard Cite

This paper aims to constitute a noteworthy contribution to the domain of direction-ofarrival (DoA) estimation through the application of deep learning algorithms. We approach the DoA estimation challenge as a binary classification task, employing a novel grid in the output layer and a deep convolutional neural network (DCNN) as the classifier. The input of the DCNN is the correlation matrix of signals received by a 4 × 4 uniformly spaced patch antenna array. The proposed model's performance is evaluated based on its capacity to predict angles of arrival from any direction in a three-dimensional space, encompassing azimuth angles within the interval [0 • , 360 • ) and polar angles within [0 • , 60 • ]. We aim to optimize the utilization of spatial information and create a robust, precise, and efficient DoA estimator. To address this, we conduct comprehensive testing in diverse scenarios, encompassing the simultaneous reception of multiple signals across a wide range of signal-to-noise ratio values. Both mean absolute error and root mean squared error are calculated to assess the performance of the DCNN. Rigorous comparison with conventional and state-of-the-art endeavors emphasizes the proposed model's efficacy.

show abstract

Distributed source DOA estimation based on deep learning networks

Tian,

Cai,

Qiu

et al. 2024

SIViP

View full text Add to dashboard Cite

Deep Learning-Based Multipath DoAs Estimation Method for mmWave Massive MIMO Systems in Low SNR

Cited by 16 publications

References 31 publications

A DOA Estimation Method Based on an Improved Transformer Model for Uniform Linear Arrays with Low SNR

A DOA Estimation Method Based on an Improved Transformer Model for Uniform Linear Arrays with Low SNR

A Novel Three-Dimensional Direction-of-Arrival Estimation Approach Using a Deep Convolutional Neural Network

Distributed source DOA estimation based on deep learning networks

Contact Info

Product

Resources

About