DROM: Optimizing the Routing in Software-Defined Networks with Deep Reinforcement Learning

Changhe Yu; Julong Lan; Zehua Guo; Yuxiang Hu

doi:10.1109/ACCESS.2018.2877686

DROM: Optimizing the Routing in Software-Defined Networks with Deep Reinforcement Learning

Changhe Yu, Julong Lan, Zehua Guo, Yuxiang Hu

Computer Science and Engineering

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

146 Scopus citations

Abstract

This paper proposes DROM, a deep reinforcement learning mechanism for Software-Defined Networks (SDN) to achieve a universal and customizable routing optimization. DROM simplifies the network operation and maintenance by improving the network performance, such as delay and throughput, with a black-box optimization in continuous time. We evaluate the DROM with experiments. The experimental results show that DROM has the good convergence and effectiveness and provides better routing configurations than existing solutions to improve the network performance, such as reducing the delay and improving the throughput.

Original language	English (US)
Article number	8502806
Pages (from-to)	64533-64539
Number of pages	7
Journal	IEEE Access
Volume	6
DOIs	https://doi.org/10.1109/ACCESS.2018.2877686
State	Published - 2018

Bibliographical note

Funding Information:
This work was supported in part by the National Natural Science Foundation of China for Innovative Research Groups under Grant 61521003, in part by the National Natural Science Foundation of China under Grant 61502530, and in part by the National High Technology Research and Development Program (863 Program) of China under Grant 2015AA016102.

Publisher Copyright:
© 2013 IEEE.

Keywords

Deep reinforcement learning
routing optimization
software-defined networking

Publisher link

10.1109/ACCESS.2018.2877686

Find It

Find It at U of M Libraries

Cite this

@article{c5effe4ea547487c9980ab5cb46566b5,

title = "DROM: Optimizing the Routing in Software-Defined Networks with Deep Reinforcement Learning",

abstract = "This paper proposes DROM, a deep reinforcement learning mechanism for Software-Defined Networks (SDN) to achieve a universal and customizable routing optimization. DROM simplifies the network operation and maintenance by improving the network performance, such as delay and throughput, with a black-box optimization in continuous time. We evaluate the DROM with experiments. The experimental results show that DROM has the good convergence and effectiveness and provides better routing configurations than existing solutions to improve the network performance, such as reducing the delay and improving the throughput.",

keywords = "Deep reinforcement learning, routing optimization, software-defined networking",

author = "Changhe Yu and Julong Lan and Zehua Guo and Yuxiang Hu",

note = "Funding Information: This work was supported in part by the National Natural Science Foundation of China for Innovative Research Groups under Grant 61521003, in part by the National Natural Science Foundation of China under Grant 61502530, and in part by the National High Technology Research and Development Program (863 Program) of China under Grant 2015AA016102. Publisher Copyright: {\textcopyright} 2013 IEEE.",

year = "2018",

doi = "10.1109/ACCESS.2018.2877686",

language = "English (US)",

volume = "6",

pages = "64533--64539",

journal = "IEEE Access",

issn = "2169-3536",

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

}

TY - JOUR

T1 - DROM

T2 - Optimizing the Routing in Software-Defined Networks with Deep Reinforcement Learning

AU - Yu, Changhe

AU - Lan, Julong

AU - Guo, Zehua

AU - Hu, Yuxiang

N1 - Funding Information: This work was supported in part by the National Natural Science Foundation of China for Innovative Research Groups under Grant 61521003, in part by the National Natural Science Foundation of China under Grant 61502530, and in part by the National High Technology Research and Development Program (863 Program) of China under Grant 2015AA016102. Publisher Copyright: © 2013 IEEE.

PY - 2018

Y1 - 2018

N2 - This paper proposes DROM, a deep reinforcement learning mechanism for Software-Defined Networks (SDN) to achieve a universal and customizable routing optimization. DROM simplifies the network operation and maintenance by improving the network performance, such as delay and throughput, with a black-box optimization in continuous time. We evaluate the DROM with experiments. The experimental results show that DROM has the good convergence and effectiveness and provides better routing configurations than existing solutions to improve the network performance, such as reducing the delay and improving the throughput.

AB - This paper proposes DROM, a deep reinforcement learning mechanism for Software-Defined Networks (SDN) to achieve a universal and customizable routing optimization. DROM simplifies the network operation and maintenance by improving the network performance, such as delay and throughput, with a black-box optimization in continuous time. We evaluate the DROM with experiments. The experimental results show that DROM has the good convergence and effectiveness and provides better routing configurations than existing solutions to improve the network performance, such as reducing the delay and improving the throughput.

KW - Deep reinforcement learning

KW - routing optimization

KW - software-defined networking

UR - http://www.scopus.com/inward/record.url?scp=85055707295&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85055707295&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2018.2877686

DO - 10.1109/ACCESS.2018.2877686

M3 - Article

AN - SCOPUS:85055707295

SN - 2169-3536

VL - 6

SP - 64533

EP - 64539

JO - IEEE Access

JF - IEEE Access

M1 - 8502806

ER -

DROM: Optimizing the Routing in Software-Defined Networks with Deep Reinforcement Learning

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this