Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels

null Narengerile; John Thompson; Paul Patras; Tharm Ratnarajah

doi:10.1109/PIMRC50174.2021.9569732

Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels

Narengerile, John Thompson, Paul Patras, Tharm Ratnarajah

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

The fifth generation wireless systems are starting to exploit the large bandwidths available in the millimeter-wave (mmWave) spectrum to provide high data rates. The exploitation of mmWave requires the use of compact antenna arrays with hundreds of antenna elements, which leads to very directional beam patterns. The beams at both the transmitter and the receiver are trained periodically to maintain accurate beam alignments. The trade-off between the training overhead and the achievable data rate must be considered. In this paper, we propose an adaptive beam training algorithm using deep reinforcement learning for tracking dynamic mmWave channels. Based on the patterns learnt from historical data, the proposed algorithm can sense the changes in the environment and switch between different beam training methods so that a high data rate can be achieved with a minimum amount of beam training.

Original language	English
Title of host publication	IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC)
Publisher	Institute of Electrical and Electronics Engineers
Pages	579-584
ISBN (Electronic)	978-1-7281-7586-7
ISBN (Print)	978-1-7281-7587-4
DOIs	https://doi.org/10.1109/PIMRC50174.2021.9569732
Publication status	Published - 13 Sept 2021
Event	2021 IEEE International Symposium on Personal, Indoor and Mobile Radio Communications - Virtual Conference Duration: 13 Sept 2021 → 16 Sept 2021 https://pimrc2021.ieee-pimrc.org/

Publication series

Name
ISSN (Print)	2166-9570
ISSN (Electronic)	2166-9589

Symposium

Symposium	2021 IEEE International Symposium on Personal, Indoor and Mobile Radio Communications
Abbreviated title	PIMRC 2021
Period	13/09/21 → 16/09/21
Internet address	https://pimrc2021.ieee-pimrc.org/

Keywords / Materials (for Non-textual outputs)

Beam training
millimeter wave
deep reinforcement learning

Access to Document

10.1109/PIMRC50174.2021.9569732Licence: All Rights Reserved

https://ieeexplore.ieee.org/document/9569732Licence: All Rights Reserved

Cite this

Narengerile, Thompson, J., Patras, P., & Ratnarajah, T. (2021). Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels. In IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC) (pp. 579-584). Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/PIMRC50174.2021.9569732

@inproceedings{0842ef17f1db464ea72ade23c79f79a7,

title = "Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels",

abstract = "The fifth generation wireless systems are starting to exploit the large bandwidths available in the millimeter-wave (mmWave) spectrum to provide high data rates. The exploitation of mmWave requires the use of compact antenna arrays with hundreds of antenna elements, which leads to very directional beam patterns. The beams at both the transmitter and the receiver are trained periodically to maintain accurate beam alignments. The trade-off between the training overhead and the achievable data rate must be considered. In this paper, we propose an adaptive beam training algorithm using deep reinforcement learning for tracking dynamic mmWave channels. Based on the patterns learnt from historical data, the proposed algorithm can sense the changes in the environment and switch between different beam training methods so that a high data rate can be achieved with a minimum amount of beam training.",

keywords = "Beam training, millimeter wave, deep reinforcement learning",

author = "Narengerile and John Thompson and Paul Patras and Tharm Ratnarajah",

year = "2021",

month = sep,

day = "13",

doi = "10.1109/PIMRC50174.2021.9569732",

language = "English",

isbn = "978-1-7281-7587-4",

publisher = "Institute of Electrical and Electronics Engineers",

pages = "579--584",

booktitle = "IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC)",

address = "United States",

note = "2021 IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2021 ; Conference date: 13-09-2021 Through 16-09-2021",

url = "https://pimrc2021.ieee-pimrc.org/",

}

Narengerile, , Thompson, J , Patras, P & Ratnarajah, T 2021, Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels. in IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC). Institute of Electrical and Electronics Engineers, pp. 579-584, 2021 IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 13/09/21. https://doi.org/10.1109/PIMRC50174.2021.9569732

Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels. / Narengerile, ; Thompson, John ; Patras, Paul et al.
IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC). Institute of Electrical and Electronics Engineers, 2021. p. 579-584.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels

AU - Narengerile, null

AU - Thompson, John

AU - Patras, Paul

AU - Ratnarajah, Tharm

PY - 2021/9/13

Y1 - 2021/9/13

N2 - The fifth generation wireless systems are starting to exploit the large bandwidths available in the millimeter-wave (mmWave) spectrum to provide high data rates. The exploitation of mmWave requires the use of compact antenna arrays with hundreds of antenna elements, which leads to very directional beam patterns. The beams at both the transmitter and the receiver are trained periodically to maintain accurate beam alignments. The trade-off between the training overhead and the achievable data rate must be considered. In this paper, we propose an adaptive beam training algorithm using deep reinforcement learning for tracking dynamic mmWave channels. Based on the patterns learnt from historical data, the proposed algorithm can sense the changes in the environment and switch between different beam training methods so that a high data rate can be achieved with a minimum amount of beam training.

AB - The fifth generation wireless systems are starting to exploit the large bandwidths available in the millimeter-wave (mmWave) spectrum to provide high data rates. The exploitation of mmWave requires the use of compact antenna arrays with hundreds of antenna elements, which leads to very directional beam patterns. The beams at both the transmitter and the receiver are trained periodically to maintain accurate beam alignments. The trade-off between the training overhead and the achievable data rate must be considered. In this paper, we propose an adaptive beam training algorithm using deep reinforcement learning for tracking dynamic mmWave channels. Based on the patterns learnt from historical data, the proposed algorithm can sense the changes in the environment and switch between different beam training methods so that a high data rate can be achieved with a minimum amount of beam training.

KW - Beam training

KW - millimeter wave

KW - deep reinforcement learning

U2 - 10.1109/PIMRC50174.2021.9569732

DO - 10.1109/PIMRC50174.2021.9569732

M3 - Conference contribution

SN - 978-1-7281-7587-4

SP - 579

EP - 584

BT - IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE PIMRC)

PB - Institute of Electrical and Electronics Engineers

T2 - 2021 IEEE International Symposium on Personal, Indoor and Mobile Radio Communications

Y2 - 13 September 2021 through 16 September 2021

ER -

Deep Reinforcement Learning-Based Beam Training for Spatially Consistent Millimeter Wave Channels

Abstract

Publication series

Symposium

Keywords / Materials (for Non-textual outputs)

Access to Document

Fingerprint

Cite this