Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach
Abstract Mobile edge computing (MEC) emerges recently as a promising solution to relieve resource-limited mobile devices from computation-intensive tasks, which enables devices to offload workloads to nearby MEC servers and improve the quality of computation experience. In this paper, an MEC enabled...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
SpringerOpen
2020-09-01
|
Series: | EURASIP Journal on Wireless Communications and Networking |
Subjects: | |
Online Access: | http://link.springer.com/article/10.1186/s13638-020-01801-6 |
id |
doaj-f54a9757e5f54c6e8dbe4231f895dfe3 |
---|---|
record_format |
Article |
spelling |
doaj-f54a9757e5f54c6e8dbe4231f895dfe32020-11-25T03:47:23ZengSpringerOpenEURASIP Journal on Wireless Communications and Networking1687-14992020-09-012020112110.1186/s13638-020-01801-6Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approachZhao Chen0Xiaodong Wang1Beijing National Research Center for Information Science and Technology, Tsinghua UniversityDepartment of Electrical Engineering, Columbia UniversityAbstract Mobile edge computing (MEC) emerges recently as a promising solution to relieve resource-limited mobile devices from computation-intensive tasks, which enables devices to offload workloads to nearby MEC servers and improve the quality of computation experience. In this paper, an MEC enabled multi-user multi-input multi-output (MIMO) system with stochastic wireless channels and task arrivals is considered. In order to minimize long-term average computation cost in terms of power consumption and buffering delay at each user, a deep reinforcement learning (DRL)-based dynamic computation offloading strategy is investigated to build a scalable system with limited feedback. Specifically, a continuous action space-based DRL approach named deep deterministic policy gradient (DDPG) is adopted to learn decentralized computation offloading policies at all users respectively, where local execution and task offloading powers will be adaptively allocated according to each user’s local observation. Numerical results demonstrate that the proposed DDPG-based strategy can help each user learn an efficient dynamic offloading policy and also verify the superiority of its continuous power allocation capability to policies learned by conventional discrete action space-based reinforcement learning approaches like deep Q-network (DQN) as well as some other greedy strategies with reduced computation cost. Besides, power-delay tradeoff for computation offloading is also analyzed for both the DDPG-based and DQN-based strategies.http://link.springer.com/article/10.1186/s13638-020-01801-6Mobile edge computingDeep reinforcement learningComputation offloadingMulti-user MIMO |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Zhao Chen Xiaodong Wang |
spellingShingle |
Zhao Chen Xiaodong Wang Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach EURASIP Journal on Wireless Communications and Networking Mobile edge computing Deep reinforcement learning Computation offloading Multi-user MIMO |
author_facet |
Zhao Chen Xiaodong Wang |
author_sort |
Zhao Chen |
title |
Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach |
title_short |
Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach |
title_full |
Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach |
title_fullStr |
Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach |
title_full_unstemmed |
Decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach |
title_sort |
decentralized computation offloading for multi-user mobile edge computing: a deep reinforcement learning approach |
publisher |
SpringerOpen |
series |
EURASIP Journal on Wireless Communications and Networking |
issn |
1687-1499 |
publishDate |
2020-09-01 |
description |
Abstract Mobile edge computing (MEC) emerges recently as a promising solution to relieve resource-limited mobile devices from computation-intensive tasks, which enables devices to offload workloads to nearby MEC servers and improve the quality of computation experience. In this paper, an MEC enabled multi-user multi-input multi-output (MIMO) system with stochastic wireless channels and task arrivals is considered. In order to minimize long-term average computation cost in terms of power consumption and buffering delay at each user, a deep reinforcement learning (DRL)-based dynamic computation offloading strategy is investigated to build a scalable system with limited feedback. Specifically, a continuous action space-based DRL approach named deep deterministic policy gradient (DDPG) is adopted to learn decentralized computation offloading policies at all users respectively, where local execution and task offloading powers will be adaptively allocated according to each user’s local observation. Numerical results demonstrate that the proposed DDPG-based strategy can help each user learn an efficient dynamic offloading policy and also verify the superiority of its continuous power allocation capability to policies learned by conventional discrete action space-based reinforcement learning approaches like deep Q-network (DQN) as well as some other greedy strategies with reduced computation cost. Besides, power-delay tradeoff for computation offloading is also analyzed for both the DDPG-based and DQN-based strategies. |
topic |
Mobile edge computing Deep reinforcement learning Computation offloading Multi-user MIMO |
url |
http://link.springer.com/article/10.1186/s13638-020-01801-6 |
work_keys_str_mv |
AT zhaochen decentralizedcomputationoffloadingformultiusermobileedgecomputingadeepreinforcementlearningapproach AT xiaodongwang decentralizedcomputationoffloadingformultiusermobileedgecomputingadeepreinforcementlearningapproach |
_version_ |
1724502068696186880 |