Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System
The rapid development of mobile device applications put tremendous pressure on edge nodes with limited computing capabilities, which may cause poor user experience. To solve this problem, collaborative cloud-edge computing is proposed. In the cloud-edge computing, an edge node with limited local res...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-08-01
|
| Series: | Frontiers in Bioengineering and Biotechnology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fbioe.2022.908056/full |
| _version_ | 1828291525815042048 |
|---|---|
| author | Jianqiao Xu Zhuohan Xu Bing Shi Bing Shi |
| author_facet | Jianqiao Xu Zhuohan Xu Bing Shi Bing Shi |
| author_sort | Jianqiao Xu |
| collection | DOAJ |
| description | The rapid development of mobile device applications put tremendous pressure on edge nodes with limited computing capabilities, which may cause poor user experience. To solve this problem, collaborative cloud-edge computing is proposed. In the cloud-edge computing, an edge node with limited local resources can rent more resources from a cloud node. According to the nature of cloud service, cloud service can be divided into private cloud and public cloud. In a private cloud environment, the edge node must allocate resources between the cloud node and the edge node. In a public cloud environment, since public cloud service providers offer various pricing modes for users’ different computing demands, the edge node also must select the appropriate pricing mode of cloud service; which is a sequential decision problem. In this stydy, we model it as a Markov decision process and parameterized action Markov decision process, and we propose a resource allocation algorithm cost efficient resource allocation with private cloud (CERAI) and cost efficient resource allocation with public cloud (CERAU) in the collaborative cloud-edge environment based on the deep reinforcement learning algorithm deep deterministic policy gradient and P-DQN. Next, we evaluated CERAI and CERAU against three typical resource allocation algorithms based on synthetic and real data of Google datasets. The experimental results demonstrate that CERAI and CERAU can effectively reduce the long-term operating cost of collaborative cloud-side computing in various demanding settings. Our analysis can provide some useful insights for enterprises to design the resource allocation strategy in the collaborative cloud-side computing system. |
| first_indexed | 2024-04-13T10:54:06Z |
| format | Article |
| id | doaj.art-1e25ec0839ab4ce188533f019c901e0d |
| institution | Directory Open Access Journal |
| issn | 2296-4185 |
| language | English |
| last_indexed | 2024-04-13T10:54:06Z |
| publishDate | 2022-08-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Bioengineering and Biotechnology |
| spelling | doaj.art-1e25ec0839ab4ce188533f019c901e0d2022-12-22T02:49:34ZengFrontiers Media S.A.Frontiers in Bioengineering and Biotechnology2296-41852022-08-011010.3389/fbioe.2022.908056908056Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing SystemJianqiao Xu0Zhuohan Xu1Bing Shi2Bing Shi3Department of Information Security, Naval University of Engineering, Wuhan, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan University of Technology, Wuhan, ChinaSchool of Computer Science and Artificial Intelligence, Wuhan University of Technology, Wuhan, ChinaShenzhen Research Institute of Wuhan University of Technology, Shenzhen, ChinaThe rapid development of mobile device applications put tremendous pressure on edge nodes with limited computing capabilities, which may cause poor user experience. To solve this problem, collaborative cloud-edge computing is proposed. In the cloud-edge computing, an edge node with limited local resources can rent more resources from a cloud node. According to the nature of cloud service, cloud service can be divided into private cloud and public cloud. In a private cloud environment, the edge node must allocate resources between the cloud node and the edge node. In a public cloud environment, since public cloud service providers offer various pricing modes for users’ different computing demands, the edge node also must select the appropriate pricing mode of cloud service; which is a sequential decision problem. In this stydy, we model it as a Markov decision process and parameterized action Markov decision process, and we propose a resource allocation algorithm cost efficient resource allocation with private cloud (CERAI) and cost efficient resource allocation with public cloud (CERAU) in the collaborative cloud-edge environment based on the deep reinforcement learning algorithm deep deterministic policy gradient and P-DQN. Next, we evaluated CERAI and CERAU against three typical resource allocation algorithms based on synthetic and real data of Google datasets. The experimental results demonstrate that CERAI and CERAU can effectively reduce the long-term operating cost of collaborative cloud-side computing in various demanding settings. Our analysis can provide some useful insights for enterprises to design the resource allocation strategy in the collaborative cloud-side computing system.https://www.frontiersin.org/articles/10.3389/fbioe.2022.908056/fullcollaborative cloud-edge computingresource allocationreinforcement learningedge computingMarkov decision process |
| spellingShingle | Jianqiao Xu Zhuohan Xu Bing Shi Bing Shi Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System Frontiers in Bioengineering and Biotechnology collaborative cloud-edge computing resource allocation reinforcement learning edge computing Markov decision process |
| title | Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System |
| title_full | Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System |
| title_fullStr | Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System |
| title_full_unstemmed | Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System |
| title_short | Deep Reinforcement Learning Based Resource Allocation Strategy in Cloud-Edge Computing System |
| title_sort | deep reinforcement learning based resource allocation strategy in cloud edge computing system |
| topic | collaborative cloud-edge computing resource allocation reinforcement learning edge computing Markov decision process |
| url | https://www.frontiersin.org/articles/10.3389/fbioe.2022.908056/full |
| work_keys_str_mv | AT jianqiaoxu deepreinforcementlearningbasedresourceallocationstrategyincloudedgecomputingsystem AT zhuohanxu deepreinforcementlearningbasedresourceallocationstrategyincloudedgecomputingsystem AT bingshi deepreinforcementlearningbasedresourceallocationstrategyincloudedgecomputingsystem AT bingshi deepreinforcementlearningbasedresourceallocationstrategyincloudedgecomputingsystem |