DQN-GNN-Based User Association Approach for Wireless Networks
In the realm of advanced mobile networks, such as the fifth generation (5G) and beyond, the increasing complexity and proliferation of devices and unique applications present a substantial challenge for User Association (UA) in wireless systems. The problem of UA in wireless networks is multifaceted...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-10-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/11/20/4286 |
| _version_ | 1797573128719695872 |
|---|---|
| author | Ibtihal Alablani Mohammed J. F. Alenazi |
| author_facet | Ibtihal Alablani Mohammed J. F. Alenazi |
| author_sort | Ibtihal Alablani |
| collection | DOAJ |
| description | In the realm of advanced mobile networks, such as the fifth generation (5G) and beyond, the increasing complexity and proliferation of devices and unique applications present a substantial challenge for User Association (UA) in wireless systems. The problem of UA in wireless networks is multifaceted and requires comprehensive exploration. This paper presents a pioneering approach to the issue, integrating a Deep Q-Network (DQN) with a Graph Neural Network (GNN) to enhance user-base station association in wireless networks. This novel approach surpasses recent methodologies, including Q-learning and max average techniques, in terms of average rewards, returns, and success rate. This superiority is attributed to its capacity to encapsulate intricate relationships and spatial dependencies among users and base stations in wireless systems. The proposed methodology achieves a success rate of 95.2%, outperforming other methodologies by a margin of up to 5.9%. |
| first_indexed | 2024-03-10T21:05:19Z |
| format | Article |
| id | doaj.art-f7391db293f1414f967a05edca760ce9 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T21:05:19Z |
| publishDate | 2023-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-f7391db293f1414f967a05edca760ce92023-11-19T17:13:50ZengMDPI AGMathematics2227-73902023-10-011120428610.3390/math11204286DQN-GNN-Based User Association Approach for Wireless NetworksIbtihal Alablani0Mohammed J. F. Alenazi1Department of Computer Engineering, College of Computer and Information Sciences, King Saud University, Riyadh P.O. Box 11451, Saudi ArabiaDepartment of Computer Engineering, College of Computer and Information Sciences, King Saud University, Riyadh P.O. Box 11451, Saudi ArabiaIn the realm of advanced mobile networks, such as the fifth generation (5G) and beyond, the increasing complexity and proliferation of devices and unique applications present a substantial challenge for User Association (UA) in wireless systems. The problem of UA in wireless networks is multifaceted and requires comprehensive exploration. This paper presents a pioneering approach to the issue, integrating a Deep Q-Network (DQN) with a Graph Neural Network (GNN) to enhance user-base station association in wireless networks. This novel approach surpasses recent methodologies, including Q-learning and max average techniques, in terms of average rewards, returns, and success rate. This superiority is attributed to its capacity to encapsulate intricate relationships and spatial dependencies among users and base stations in wireless systems. The proposed methodology achieves a success rate of 95.2%, outperforming other methodologies by a margin of up to 5.9%.https://www.mdpi.com/2227-7390/11/20/4286Graph Neural NetworksDeep Q-NetworkUser Association5GMachine LearningReinforcement Learning |
| spellingShingle | Ibtihal Alablani Mohammed J. F. Alenazi DQN-GNN-Based User Association Approach for Wireless Networks Mathematics Graph Neural Networks Deep Q-Network User Association 5G Machine Learning Reinforcement Learning |
| title | DQN-GNN-Based User Association Approach for Wireless Networks |
| title_full | DQN-GNN-Based User Association Approach for Wireless Networks |
| title_fullStr | DQN-GNN-Based User Association Approach for Wireless Networks |
| title_full_unstemmed | DQN-GNN-Based User Association Approach for Wireless Networks |
| title_short | DQN-GNN-Based User Association Approach for Wireless Networks |
| title_sort | dqn gnn based user association approach for wireless networks |
| topic | Graph Neural Networks Deep Q-Network User Association 5G Machine Learning Reinforcement Learning |
| url | https://www.mdpi.com/2227-7390/11/20/4286 |
| work_keys_str_mv | AT ibtihalalablani dqngnnbaseduserassociationapproachforwirelessnetworks AT mohammedjfalenazi dqngnnbaseduserassociationapproachforwirelessnetworks |