Attention-based reinforcement learning for real-time UAV semantic communication

Won Joon Yun; Byungju Lim; Soyi Jung; Young Chai Ko; Jihong Park; Joongheon Kim; Mehdi Bennis

doi:10.1109/ISWCS49558.2021.9562230

Attention-based reinforcement learning for real-time UAV semantic communication

Won Joon Yun, Byungju Lim, Soyi Jung, Young Chai Ko, Jihong Park, Joongheon Kim, Mehdi Bennis

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

31 Citations (Scopus)

Abstract

In this article, we study the problem of air-to-ground ultra-reliable and low-latency communication (URLLC) for a moving ground user. This is done by controlling multiple unmanned aerial vehicles (UAVs) in real time while avoiding inter-UAV collisions. To this end, we propose a novel multiagent deep reinforcement learning (MADRL) framework, coined a graph attention exchange network (GAXNet). In GAXNet, each UAV constructs an attention graph locally measuring the level of attention to its neighboring UAVs, while exchanging the attention weights with other UAVs so as to reduce the attention mismatch between them. Simulation results corroborates that GAXNet achieves up to 4.5x higher rewards during training. At execution, without incurring inter-UAV collisions, G2ANet improves reliability of air-to-ground network in terms of latency and error rate.

Original language	English
Title of host publication	2021 17th International Symposium on Wireless Communication Systems, ISWCS 2021
Publisher	VDE Verlag GmbH
ISBN (Electronic)	9781728174327
DOIs	https://doi.org/10.1109/ISWCS49558.2021.9562230
Publication status	Published - 2021 Sept 6
Event	17th International Symposium on Wireless Communication Systems, ISWCS 2021 - Berlin, Germany Duration: 2021 Sept 6 → 2021 Sept 9

Publication series

Name	Proceedings of the International Symposium on Wireless Communication Systems
Volume	2021-September
ISSN (Print)	2154-0217
ISSN (Electronic)	2154-0225

Conference

Conference	17th International Symposium on Wireless Communication Systems, ISWCS 2021
Country/Territory	Germany
City	Berlin
Period	21/9/6 → 21/9/9

Bibliographical note

Publisher Copyright:
© 2021 IEEE

ASJC Scopus subject areas

Computer Networks and Communications
Electrical and Electronic Engineering
Communication

Access to Document

10.1109/ISWCS49558.2021.9562230

Cite this

Yun, W. J., Lim, B., Jung, S., Ko, Y. C., Park, J., Kim, J., & Bennis, M. (2021). Attention-based reinforcement learning for real-time UAV semantic communication. In 2021 17th International Symposium on Wireless Communication Systems, ISWCS 2021 (Proceedings of the International Symposium on Wireless Communication Systems; Vol. 2021-September). VDE Verlag GmbH. https://doi.org/10.1109/ISWCS49558.2021.9562230

Attention-based reinforcement learning for real-time UAV semantic communication. / Yun, Won Joon; Lim, Byungju; Jung, Soyi et al.
2021 17th International Symposium on Wireless Communication Systems, ISWCS 2021. VDE Verlag GmbH, 2021. (Proceedings of the International Symposium on Wireless Communication Systems; Vol. 2021-September).

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

Yun, WJ, Lim, B, Jung, S, Ko, YC, Park, J, Kim, J & Bennis, M 2021, Attention-based reinforcement learning for real-time UAV semantic communication. in 2021 17th International Symposium on Wireless Communication Systems, ISWCS 2021. Proceedings of the International Symposium on Wireless Communication Systems, vol. 2021-September, VDE Verlag GmbH, 17th International Symposium on Wireless Communication Systems, ISWCS 2021, Berlin, Germany, 21/9/6. https://doi.org/10.1109/ISWCS49558.2021.9562230

@inproceedings{82b554682ab94d629e8567ad2c107fa9,

title = "Attention-based reinforcement learning for real-time UAV semantic communication",

abstract = "In this article, we study the problem of air-to-ground ultra-reliable and low-latency communication (URLLC) for a moving ground user. This is done by controlling multiple unmanned aerial vehicles (UAVs) in real time while avoiding inter-UAV collisions. To this end, we propose a novel multiagent deep reinforcement learning (MADRL) framework, coined a graph attention exchange network (GAXNet). In GAXNet, each UAV constructs an attention graph locally measuring the level of attention to its neighboring UAVs, while exchanging the attention weights with other UAVs so as to reduce the attention mismatch between them. Simulation results corroborates that GAXNet achieves up to 4.5x higher rewards during training. At execution, without incurring inter-UAV collisions, G2ANet improves reliability of air-to-ground network in terms of latency and error rate.",

author = "Yun, {Won Joon} and Byungju Lim and Soyi Jung and Ko, {Young Chai} and Jihong Park and Joongheon Kim and Mehdi Bennis",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE; 17th International Symposium on Wireless Communication Systems, ISWCS 2021 ; Conference date: 06-09-2021 Through 09-09-2021",

year = "2021",

month = sep,

day = "6",

doi = "10.1109/ISWCS49558.2021.9562230",

language = "English",

series = "Proceedings of the International Symposium on Wireless Communication Systems",

publisher = "VDE Verlag GmbH",

booktitle = "2021 17th International Symposium on Wireless Communication Systems, ISWCS 2021",

address = "Germany",

}

TY - GEN

T1 - Attention-based reinforcement learning for real-time UAV semantic communication

AU - Yun, Won Joon

AU - Lim, Byungju

AU - Jung, Soyi

AU - Ko, Young Chai

AU - Park, Jihong

AU - Kim, Joongheon

AU - Bennis, Mehdi

PY - 2021/9/6

Y1 - 2021/9/6

N2 - In this article, we study the problem of air-to-ground ultra-reliable and low-latency communication (URLLC) for a moving ground user. This is done by controlling multiple unmanned aerial vehicles (UAVs) in real time while avoiding inter-UAV collisions. To this end, we propose a novel multiagent deep reinforcement learning (MADRL) framework, coined a graph attention exchange network (GAXNet). In GAXNet, each UAV constructs an attention graph locally measuring the level of attention to its neighboring UAVs, while exchanging the attention weights with other UAVs so as to reduce the attention mismatch between them. Simulation results corroborates that GAXNet achieves up to 4.5x higher rewards during training. At execution, without incurring inter-UAV collisions, G2ANet improves reliability of air-to-ground network in terms of latency and error rate.

AB - In this article, we study the problem of air-to-ground ultra-reliable and low-latency communication (URLLC) for a moving ground user. This is done by controlling multiple unmanned aerial vehicles (UAVs) in real time while avoiding inter-UAV collisions. To this end, we propose a novel multiagent deep reinforcement learning (MADRL) framework, coined a graph attention exchange network (GAXNet). In GAXNet, each UAV constructs an attention graph locally measuring the level of attention to its neighboring UAVs, while exchanging the attention weights with other UAVs so as to reduce the attention mismatch between them. Simulation results corroborates that GAXNet achieves up to 4.5x higher rewards during training. At execution, without incurring inter-UAV collisions, G2ANet improves reliability of air-to-ground network in terms of latency and error rate.

UR - http://www.scopus.com/inward/record.url?scp=85118152186&partnerID=8YFLogxK

U2 - 10.1109/ISWCS49558.2021.9562230

DO - 10.1109/ISWCS49558.2021.9562230

M3 - Conference contribution

AN - SCOPUS:85118152186

T3 - Proceedings of the International Symposium on Wireless Communication Systems

BT - 2021 17th International Symposium on Wireless Communication Systems, ISWCS 2021

PB - VDE Verlag GmbH

T2 - 17th International Symposium on Wireless Communication Systems, ISWCS 2021

Y2 - 6 September 2021 through 9 September 2021

ER -

Attention-based reinforcement learning for real-time UAV semantic communication

Abstract

Publication series

Conference

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this