DQN-Based adaptive modulation scheme over wireless communication channels

Donggu Lee; Young Ghyu Sun; Soo Hyun Kim; Isaac Sim; Yu Min Hwang; Yoan Shin; Dong In Kim; Jin Young Kim

doi:10.1109/LCOMM.2020.2978390

DQN-Based adaptive modulation scheme over wireless communication channels

Donggu Lee
, Young Ghyu Sun
, Soo Hyun Kim
, Isaac Sim
, Yu Min Hwang
, Yoan Shin
, Dong In Kim
, Jin Young Kim

Research output: Contribution to journal › Article › peer-review

35 Scopus citations

Abstract

In this letter, to improve data rate over wireless communication channels, we propose a deep Q network (DQN)-based adaptive modulation scheme by using Markov decision process (MDP) model. The proposed algorithm makes the reinforcement learning agent to select rate region boundaries as the states, which divide signal-to-noise ratio (SNR) range into rate regions. The simulation results show that spectral efficiency can be improved on the average by 0.5395 bps/Hz in wide SNR range.

Original language	English
Article number	9025237
Pages (from-to)	1289-1293
Number of pages	5
Journal	IEEE Communications Letters
Volume	24
Issue number	6
DOIs	https://doi.org/10.1109/LCOMM.2020.2978390
State	Published - Jun 2020
Externally published	Yes

Keywords

Adaptive modulation
deep learning
deep Q network
reinforcement learning

Access to Document

10.1109/LCOMM.2020.2978390

Cite this

@article{b574bf2f46074079a854df148bea1f0e,

title = "DQN-Based adaptive modulation scheme over wireless communication channels",

abstract = "In this letter, to improve data rate over wireless communication channels, we propose a deep Q network (DQN)-based adaptive modulation scheme by using Markov decision process (MDP) model. The proposed algorithm makes the reinforcement learning agent to select rate region boundaries as the states, which divide signal-to-noise ratio (SNR) range into rate regions. The simulation results show that spectral efficiency can be improved on the average by 0.5395 bps/Hz in wide SNR range.",

keywords = "Adaptive modulation, deep learning, deep Q network, reinforcement learning",

author = "Donggu Lee and Sun, \{Young Ghyu\} and Kim, \{Soo Hyun\} and Isaac Sim and Hwang, \{Yu Min\} and Yoan Shin and Kim, \{Dong In\} and Kim, \{Jin Young\}",

note = "Publisher Copyright: {\textcopyright} 1997-2012 IEEE.",

year = "2020",

month = jun,

doi = "10.1109/LCOMM.2020.2978390",

language = "English",

volume = "24",

pages = "1289--1293",

journal = "IEEE Communications Letters",

issn = "1089-7798",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "6",

}

TY - JOUR

T1 - DQN-Based adaptive modulation scheme over wireless communication channels

AU - Lee, Donggu

AU - Sun, Young Ghyu

AU - Kim, Soo Hyun

AU - Sim, Isaac

AU - Hwang, Yu Min

AU - Shin, Yoan

AU - Kim, Dong In

AU - Kim, Jin Young

PY - 2020/6

Y1 - 2020/6

N2 - In this letter, to improve data rate over wireless communication channels, we propose a deep Q network (DQN)-based adaptive modulation scheme by using Markov decision process (MDP) model. The proposed algorithm makes the reinforcement learning agent to select rate region boundaries as the states, which divide signal-to-noise ratio (SNR) range into rate regions. The simulation results show that spectral efficiency can be improved on the average by 0.5395 bps/Hz in wide SNR range.

AB - In this letter, to improve data rate over wireless communication channels, we propose a deep Q network (DQN)-based adaptive modulation scheme by using Markov decision process (MDP) model. The proposed algorithm makes the reinforcement learning agent to select rate region boundaries as the states, which divide signal-to-noise ratio (SNR) range into rate regions. The simulation results show that spectral efficiency can be improved on the average by 0.5395 bps/Hz in wide SNR range.

KW - Adaptive modulation

KW - deep learning

KW - deep Q network

KW - reinforcement learning

UR - https://www.scopus.com/pages/publications/85086470333

U2 - 10.1109/LCOMM.2020.2978390

DO - 10.1109/LCOMM.2020.2978390

M3 - Article

AN - SCOPUS:85086470333

SN - 1089-7798

VL - 24

SP - 1289

EP - 1293

JO - IEEE Communications Letters

JF - IEEE Communications Letters

IS - 6

M1 - 9025237

ER -

DQN-Based adaptive modulation scheme over wireless communication channels

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this