Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning
Abstract Intelligent reflective surface (IRS) provides an effective solution for reconfiguring air‐to‐ground wireless channels, and intelligent agents based on reinforcement learning can dynamically adjust the reflection coefficient of IRS to adapt to changing channels. However, most exiting IRS con...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-02-01
|
| Series: | IET Communications |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/cmu2.12718 |
| _version_ | 1797311448668438528 |
|---|---|
| author | Yifan Du Nan Qi Kewei Wang Ming Xiao Wenjing Wang |
| author_facet | Yifan Du Nan Qi Kewei Wang Ming Xiao Wenjing Wang |
| author_sort | Yifan Du |
| collection | DOAJ |
| description | Abstract Intelligent reflective surface (IRS) provides an effective solution for reconfiguring air‐to‐ground wireless channels, and intelligent agents based on reinforcement learning can dynamically adjust the reflection coefficient of IRS to adapt to changing channels. However, most exiting IRS configuration schemes based on reinforcement learning require long training time and are difficult to be industrially deployed. This paper, proposes a model‐free IRS control scheme based on reinforcement learning and adopts transfer learning to accelerate the training process. A knowledge base of the source tasks has been constructed for transfer learning, allowing accumulation of experience from different source tasks. To mitigate potential negative effects of transfer learning, quantitative analysis of task similarity through unmanned aerial vehicle (UAV) flight path is conducted. After identifying the most similar source task to the target task, parameters of the source task model are used as the initial values for the target task model to accelerate the convergence process of reinforcement learning. Simulation results demonstrate that the proposed method can increase the convergence speed of the traditional DDQN algorithm by up to 60%. |
| first_indexed | 2024-03-08T01:58:52Z |
| format | Article |
| id | doaj.art-6f9fff38dfcc4c1cad2674f0528da430 |
| institution | Directory Open Access Journal |
| issn | 1751-8628 1751-8636 |
| language | English |
| last_indexed | 2024-03-08T01:58:52Z |
| publishDate | 2024-02-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Communications |
| spelling | doaj.art-6f9fff38dfcc4c1cad2674f0528da4302024-02-14T06:55:54ZengWileyIET Communications1751-86281751-86362024-02-0118321422410.1049/cmu2.12718Intelligent reflecting surface‐assisted UAV inspection system based on transfer learningYifan Du0Nan Qi1Kewei Wang2Ming Xiao3Wenjing Wang4Department of Electrical Engineering Nanjing University of Aeronautics and Astronautics Nanjing ChinaDepartment of Electrical Engineering Nanjing University of Aeronautics and Astronautics Nanjing ChinaDepartment of Electrical Engineering Nanjing University of Aeronautics and Astronautics Nanjing ChinaSchool of Electrical Engineering of KTH Royal Institute of Technology Stockholm SwedenSchool of Communication and Information Engineering Xi'an University of Posts and Telecommunications Xi'an ChinaAbstract Intelligent reflective surface (IRS) provides an effective solution for reconfiguring air‐to‐ground wireless channels, and intelligent agents based on reinforcement learning can dynamically adjust the reflection coefficient of IRS to adapt to changing channels. However, most exiting IRS configuration schemes based on reinforcement learning require long training time and are difficult to be industrially deployed. This paper, proposes a model‐free IRS control scheme based on reinforcement learning and adopts transfer learning to accelerate the training process. A knowledge base of the source tasks has been constructed for transfer learning, allowing accumulation of experience from different source tasks. To mitigate potential negative effects of transfer learning, quantitative analysis of task similarity through unmanned aerial vehicle (UAV) flight path is conducted. After identifying the most similar source task to the target task, parameters of the source task model are used as the initial values for the target task model to accelerate the convergence process of reinforcement learning. Simulation results demonstrate that the proposed method can increase the convergence speed of the traditional DDQN algorithm by up to 60%.https://doi.org/10.1049/cmu2.127186Glearning (artificial intelligence) |
| spellingShingle | Yifan Du Nan Qi Kewei Wang Ming Xiao Wenjing Wang Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning IET Communications 6G learning (artificial intelligence) |
| title | Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning |
| title_full | Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning |
| title_fullStr | Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning |
| title_full_unstemmed | Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning |
| title_short | Intelligent reflecting surface‐assisted UAV inspection system based on transfer learning |
| title_sort | intelligent reflecting surface assisted uav inspection system based on transfer learning |
| topic | 6G learning (artificial intelligence) |
| url | https://doi.org/10.1049/cmu2.12718 |
| work_keys_str_mv | AT yifandu intelligentreflectingsurfaceassisteduavinspectionsystembasedontransferlearning AT nanqi intelligentreflectingsurfaceassisteduavinspectionsystembasedontransferlearning AT keweiwang intelligentreflectingsurfaceassisteduavinspectionsystembasedontransferlearning AT mingxiao intelligentreflectingsurfaceassisteduavinspectionsystembasedontransferlearning AT wenjingwang intelligentreflectingsurfaceassisteduavinspectionsystembasedontransferlearning |