Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing
Using reinforcement learning technologies to learn offloading strategies for multi-access edge computing systems has been developed by researchers. However, large-scale systems are unsuitable for reinforcement learning, due to their huge state spaces and offloading behaviors. For this reason, this w...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-08-01
|
Series: | Information |
Subjects: | |
Online Access: | https://www.mdpi.com/2078-2489/12/9/343 |
id |
doaj-65845501de21434f87e89bbdc54b17b5 |
---|---|
record_format |
Article |
spelling |
doaj-65845501de21434f87e89bbdc54b17b52021-09-26T00:26:23ZengMDPI AGInformation2078-24892021-08-011234334310.3390/info12090343Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge ComputingChunyang Hu0Jingchen Li1Haobin Shi2Bin Ning3Qiong Gu4Hubei Key Laboratory of Power System Design and Test for Electrical Vehicle, Hubei University of Arts and Science, Xiangyang 441053, ChinaSchool of Computer Science and Engineering, Northwestern Polytechnical University, Xi’an 710129, ChinaSchool of Computer Science and Engineering, Northwestern Polytechnical University, Xi’an 710129, ChinaSchool of Computer Engineering, Hubei University of Arts and Science, Xiangyang 441053, ChinaSchool of Computer Engineering, Hubei University of Arts and Science, Xiangyang 441053, ChinaUsing reinforcement learning technologies to learn offloading strategies for multi-access edge computing systems has been developed by researchers. However, large-scale systems are unsuitable for reinforcement learning, due to their huge state spaces and offloading behaviors. For this reason, this work introduces the centralized training and decentralized execution mechanism, designing a decentralized reinforcement learning model for multi-access edge computing systems. Considering a cloud server and several edge servers, we separate the training and execution in the reinforcement learning model. The execution happens in edge devices of the system, and edge servers need no communication. Conversely, the training process occurs at the cloud device, which causes a lower transmission latency. The developed method uses a deep deterministic policy gradient algorithm to optimize offloading strategies. The simulated experiment shows that our method can learn the offloading strategy for each edge device efficiently.https://www.mdpi.com/2078-2489/12/9/343multi-access edge computingdeep reinforcement learningtask offloading |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Chunyang Hu Jingchen Li Haobin Shi Bin Ning Qiong Gu |
spellingShingle |
Chunyang Hu Jingchen Li Haobin Shi Bin Ning Qiong Gu Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing Information multi-access edge computing deep reinforcement learning task offloading |
author_facet |
Chunyang Hu Jingchen Li Haobin Shi Bin Ning Qiong Gu |
author_sort |
Chunyang Hu |
title |
Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing |
title_short |
Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing |
title_full |
Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing |
title_fullStr |
Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing |
title_full_unstemmed |
Decentralized Offloading Strategies Based on Reinforcement Learning for Multi-Access Edge Computing |
title_sort |
decentralized offloading strategies based on reinforcement learning for multi-access edge computing |
publisher |
MDPI AG |
series |
Information |
issn |
2078-2489 |
publishDate |
2021-08-01 |
description |
Using reinforcement learning technologies to learn offloading strategies for multi-access edge computing systems has been developed by researchers. However, large-scale systems are unsuitable for reinforcement learning, due to their huge state spaces and offloading behaviors. For this reason, this work introduces the centralized training and decentralized execution mechanism, designing a decentralized reinforcement learning model for multi-access edge computing systems. Considering a cloud server and several edge servers, we separate the training and execution in the reinforcement learning model. The execution happens in edge devices of the system, and edge servers need no communication. Conversely, the training process occurs at the cloud device, which causes a lower transmission latency. The developed method uses a deep deterministic policy gradient algorithm to optimize offloading strategies. The simulated experiment shows that our method can learn the offloading strategy for each edge device efficiently. |
topic |
multi-access edge computing deep reinforcement learning task offloading |
url |
https://www.mdpi.com/2078-2489/12/9/343 |
work_keys_str_mv |
AT chunyanghu decentralizedoffloadingstrategiesbasedonreinforcementlearningformultiaccessedgecomputing AT jingchenli decentralizedoffloadingstrategiesbasedonreinforcementlearningformultiaccessedgecomputing AT haobinshi decentralizedoffloadingstrategiesbasedonreinforcementlearningformultiaccessedgecomputing AT binning decentralizedoffloadingstrategiesbasedonreinforcementlearningformultiaccessedgecomputing AT qionggu decentralizedoffloadingstrategiesbasedonreinforcementlearningformultiaccessedgecomputing |
_version_ |
1717366155924996096 |