Full Text Views
92
Direction of arrival (DOA) estimation of wireless signals is demanded in many applications. In addition to classical methods such as MUSIC and ESPRIT, non-linear algorithms such as compressed sensing have become common subjects of study recently. Deep learning or machine learning is also known as a non-linear algorithm and has been applied in various fields. Generally, DOA estimation using deep learning is classified as on-grid estimation. A major problem of on-grid estimation is that the accuracy may be degraded when the DOA is near the boundary. To reduce such estimation errors, we propose a method of combining two DNNs whose grids are offset by one half of the grid size. Simulation results show that our proposal outperforms MUSIC which is a typical off-grid estimation method. Furthermore, it is shown that the DNN specially trained for a close DOA case achieves very high accuracy for that case compared with MUSIC.
Yuya KASE
Hokkaido University
Toshihiko NISHIMURA
Hokkaido University
Takeo OHGANE
Hokkaido University
Yasutaka OGAWA
Hokkaido University
Takanori SATO
Hokkaido University
Yoshihisa KISHIYAMA
NTT DOCOMO, INC.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Yuya KASE, Toshihiko NISHIMURA, Takeo OHGANE, Yasutaka OGAWA, Takanori SATO, Yoshihisa KISHIYAMA, "Accuracy Improvement in DOA Estimation with Deep Learning" in IEICE TRANSACTIONS on Communications,
vol. E105-B, no. 5, pp. 588-599, May 2022, doi: 10.1587/transcom.2021EBT0001.
Abstract: Direction of arrival (DOA) estimation of wireless signals is demanded in many applications. In addition to classical methods such as MUSIC and ESPRIT, non-linear algorithms such as compressed sensing have become common subjects of study recently. Deep learning or machine learning is also known as a non-linear algorithm and has been applied in various fields. Generally, DOA estimation using deep learning is classified as on-grid estimation. A major problem of on-grid estimation is that the accuracy may be degraded when the DOA is near the boundary. To reduce such estimation errors, we propose a method of combining two DNNs whose grids are offset by one half of the grid size. Simulation results show that our proposal outperforms MUSIC which is a typical off-grid estimation method. Furthermore, it is shown that the DNN specially trained for a close DOA case achieves very high accuracy for that case compared with MUSIC.
URL: https://global.ieice.org/en_transactions/communications/10.1587/transcom.2021EBT0001/_p
Copy
@ARTICLE{e105-b_5_588,
author={Yuya KASE, Toshihiko NISHIMURA, Takeo OHGANE, Yasutaka OGAWA, Takanori SATO, Yoshihisa KISHIYAMA, },
journal={IEICE TRANSACTIONS on Communications},
title={Accuracy Improvement in DOA Estimation with Deep Learning},
year={2022},
volume={E105-B},
number={5},
pages={588-599},
abstract={Direction of arrival (DOA) estimation of wireless signals is demanded in many applications. In addition to classical methods such as MUSIC and ESPRIT, non-linear algorithms such as compressed sensing have become common subjects of study recently. Deep learning or machine learning is also known as a non-linear algorithm and has been applied in various fields. Generally, DOA estimation using deep learning is classified as on-grid estimation. A major problem of on-grid estimation is that the accuracy may be degraded when the DOA is near the boundary. To reduce such estimation errors, we propose a method of combining two DNNs whose grids are offset by one half of the grid size. Simulation results show that our proposal outperforms MUSIC which is a typical off-grid estimation method. Furthermore, it is shown that the DNN specially trained for a close DOA case achieves very high accuracy for that case compared with MUSIC.},
keywords={},
doi={10.1587/transcom.2021EBT0001},
ISSN={1745-1345},
month={May},}
Copy
TY - JOUR
TI - Accuracy Improvement in DOA Estimation with Deep Learning
T2 - IEICE TRANSACTIONS on Communications
SP - 588
EP - 599
AU - Yuya KASE
AU - Toshihiko NISHIMURA
AU - Takeo OHGANE
AU - Yasutaka OGAWA
AU - Takanori SATO
AU - Yoshihisa KISHIYAMA
PY - 2022
DO - 10.1587/transcom.2021EBT0001
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E105-B
IS - 5
JA - IEICE TRANSACTIONS on Communications
Y1 - May 2022
AB - Direction of arrival (DOA) estimation of wireless signals is demanded in many applications. In addition to classical methods such as MUSIC and ESPRIT, non-linear algorithms such as compressed sensing have become common subjects of study recently. Deep learning or machine learning is also known as a non-linear algorithm and has been applied in various fields. Generally, DOA estimation using deep learning is classified as on-grid estimation. A major problem of on-grid estimation is that the accuracy may be degraded when the DOA is near the boundary. To reduce such estimation errors, we propose a method of combining two DNNs whose grids are offset by one half of the grid size. Simulation results show that our proposal outperforms MUSIC which is a typical off-grid estimation method. Furthermore, it is shown that the DNN specially trained for a close DOA case achieves very high accuracy for that case compared with MUSIC.
ER -