Deep Reinforcement Learning for Resource Management in Network Slicing

Rongpeng Li, Zhifeng Zhao, Qi Sun, I. Chi-Lin, Chenyang Yang, Xianfu Chen, Minjian Zhao, Honggang Zhang

Research output: Contribution to journalArticleScientificpeer-review

5 Citations (Scopus)

Abstract

Network slicing is born as an emerging business to operators, by allowing them to sell the customized slices to various tenants at different prices. In order to provide better-performing and cost-efficient services, network slicing involves challenging technical issues and urgently looks forward to intelligent innovations to make the resource management consistent with users’ activities per slice. In that regard, deep reinforcement learning (DRL), which focuses on how to interact with the environment by trying alternative actions and reinforcing the tendency actions producing more rewarding consequences, is assumed to be a promising solution. In this paper, after briefly reviewing the fundamental concepts of DRL, we investigate the application of DRL in solving some typical resource management for network slicing scenarios, which include radio resource slicing and priority-based core network slicing, and demonstrate the advantage of DRL over several competing schemes through extensive simulations. Finally, we also discuss the possible challenges to apply DRL in network slicing from a general perspective.
Original languageEnglish
Article number8540003
Pages (from-to)74429 - 74441
Number of pages13
JournalIEEE Access
Volume6
DOIs
Publication statusPublished - 2018
MoE publication typeNot Eligible

Fingerprint

Reinforcement learning
Innovation
Costs
Industry

Keywords

  • 5G mobile communication
  • Deep Reinforcement Learning
  • Network slicing
  • Network Slicing
  • Neural networks
  • Neural Networks
  • Q-Learning
  • Quality of experience
  • Resource management
  • Resource Management

Cite this

Li, R., Zhao, Z., Sun, Q., Chi-Lin, I., Yang, C., Chen, X., ... Zhang, H. (2018). Deep Reinforcement Learning for Resource Management in Network Slicing. IEEE Access, 6, 74429 - 74441. [8540003]. https://doi.org/10.1109/ACCESS.2018.2881964
Li, Rongpeng ; Zhao, Zhifeng ; Sun, Qi ; Chi-Lin, I. ; Yang, Chenyang ; Chen, Xianfu ; Zhao, Minjian ; Zhang, Honggang. / Deep Reinforcement Learning for Resource Management in Network Slicing. In: IEEE Access. 2018 ; Vol. 6. pp. 74429 - 74441.
@article{e2f88268fb4a44c68ac135d465e0d64b,
title = "Deep Reinforcement Learning for Resource Management in Network Slicing",
abstract = "Network slicing is born as an emerging business to operators, by allowing them to sell the customized slices to various tenants at different prices. In order to provide better-performing and cost-efficient services, network slicing involves challenging technical issues and urgently looks forward to intelligent innovations to make the resource management consistent with users’ activities per slice. In that regard, deep reinforcement learning (DRL), which focuses on how to interact with the environment by trying alternative actions and reinforcing the tendency actions producing more rewarding consequences, is assumed to be a promising solution. In this paper, after briefly reviewing the fundamental concepts of DRL, we investigate the application of DRL in solving some typical resource management for network slicing scenarios, which include radio resource slicing and priority-based core network slicing, and demonstrate the advantage of DRL over several competing schemes through extensive simulations. Finally, we also discuss the possible challenges to apply DRL in network slicing from a general perspective.",
keywords = "5G mobile communication, Deep Reinforcement Learning, Network slicing, Network Slicing, Neural networks, Neural Networks, Q-Learning, Quality of experience, Resource management, Resource Management",
author = "Rongpeng Li and Zhifeng Zhao and Qi Sun and I. Chi-Lin and Chenyang Yang and Xianfu Chen and Minjian Zhao and Honggang Zhang",
year = "2018",
doi = "10.1109/ACCESS.2018.2881964",
language = "English",
volume = "6",
pages = "74429 -- 74441",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "IEEE Institute of Electrical and Electronic Engineers",

}

Li, R, Zhao, Z, Sun, Q, Chi-Lin, I, Yang, C, Chen, X, Zhao, M & Zhang, H 2018, 'Deep Reinforcement Learning for Resource Management in Network Slicing', IEEE Access, vol. 6, 8540003, pp. 74429 - 74441. https://doi.org/10.1109/ACCESS.2018.2881964

Deep Reinforcement Learning for Resource Management in Network Slicing. / Li, Rongpeng; Zhao, Zhifeng; Sun, Qi; Chi-Lin, I.; Yang, Chenyang; Chen, Xianfu; Zhao, Minjian; Zhang, Honggang.

In: IEEE Access, Vol. 6, 8540003, 2018, p. 74429 - 74441.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Deep Reinforcement Learning for Resource Management in Network Slicing

AU - Li, Rongpeng

AU - Zhao, Zhifeng

AU - Sun, Qi

AU - Chi-Lin, I.

AU - Yang, Chenyang

AU - Chen, Xianfu

AU - Zhao, Minjian

AU - Zhang, Honggang

PY - 2018

Y1 - 2018

N2 - Network slicing is born as an emerging business to operators, by allowing them to sell the customized slices to various tenants at different prices. In order to provide better-performing and cost-efficient services, network slicing involves challenging technical issues and urgently looks forward to intelligent innovations to make the resource management consistent with users’ activities per slice. In that regard, deep reinforcement learning (DRL), which focuses on how to interact with the environment by trying alternative actions and reinforcing the tendency actions producing more rewarding consequences, is assumed to be a promising solution. In this paper, after briefly reviewing the fundamental concepts of DRL, we investigate the application of DRL in solving some typical resource management for network slicing scenarios, which include radio resource slicing and priority-based core network slicing, and demonstrate the advantage of DRL over several competing schemes through extensive simulations. Finally, we also discuss the possible challenges to apply DRL in network slicing from a general perspective.

AB - Network slicing is born as an emerging business to operators, by allowing them to sell the customized slices to various tenants at different prices. In order to provide better-performing and cost-efficient services, network slicing involves challenging technical issues and urgently looks forward to intelligent innovations to make the resource management consistent with users’ activities per slice. In that regard, deep reinforcement learning (DRL), which focuses on how to interact with the environment by trying alternative actions and reinforcing the tendency actions producing more rewarding consequences, is assumed to be a promising solution. In this paper, after briefly reviewing the fundamental concepts of DRL, we investigate the application of DRL in solving some typical resource management for network slicing scenarios, which include radio resource slicing and priority-based core network slicing, and demonstrate the advantage of DRL over several competing schemes through extensive simulations. Finally, we also discuss the possible challenges to apply DRL in network slicing from a general perspective.

KW - 5G mobile communication

KW - Deep Reinforcement Learning

KW - Network slicing

KW - Network Slicing

KW - Neural networks

KW - Neural Networks

KW - Q-Learning

KW - Quality of experience

KW - Resource management

KW - Resource Management

UR - http://www.scopus.com/inward/record.url?scp=85056721900&partnerID=8YFLogxK

U2 - 10.1109/ACCESS.2018.2881964

DO - 10.1109/ACCESS.2018.2881964

M3 - Article

AN - SCOPUS:85056721900

VL - 6

SP - 74429

EP - 74441

JO - IEEE Access

JF - IEEE Access

SN - 2169-3536

M1 - 8540003

ER -