Direction of Arrival Estimation in Terahertz Communications using Convolutional Neural Networks

Date

2024

Authors

Abdullah, M.
Li, M.S.
He, J.
Wang, K.
Fumeaux, C.
Withayachumnankul, W.

Editors

Advisors

Journal Title

Journal ISSN

Volume Title

Type:

Conference paper

Citation

International Conference on Infrared, Millimeter, and Terahertz Waves, IRMMW-THz, 2024, pp.1-2

Statement of Responsibility

Mariam Abdullah, Mingxiang Stephen Li, Jiayuan He, Ke Wang, Christophe Fumeaux, and Withawat Withayachumnankul

Conference Name

International Conference on Infrared, Millimeter, and Terahertz Waves, IRMMW-THz (1 Sep 2024 - 6 Sep 2024 : Perth, Western Australia)

Abstract

We demonstrate an approach to accurately estimating the Direction of Arrival (DoA) in terahertz communications using Convolutional Neural Networks (CNNs). Quasi-random patterns are generated with a frequency-diverse antenna which is deliberately designed to break symmetry, and a CNN model is then trained to capture the relationship between the spectrally resolved radiation patterns and their respective angles of arrival. The CNN converges to a minimum validation mean squared error (MSE) of 3.9◦ and root mean squared error (RMSE) of 1.9◦ .

School/Discipline

Dissertation Note

Provenance

Description

Access Status

Rights

© 2024 IEEE

License

Grant ID

Call number

Persistent link to this record