| 大会名称 |
| 2021年 総合大会 |
| 大会コ－ド |
| 2021G |
| 開催年 |
|
2021
|
| 発行日 |
|
2021-02-23
|
| セッション番号 |
|
BS-7
|
| セッション名 |
|
AI technologies and their applications for future network systems and services
|
| 講演日 |
|
2021/3/9
|
| 講演場所(会議室等) |
|
Meeting 30
|
| 講演番号 |
|
BS-7-9
|
| タイトル |
|
Q-learning Based Path Planning for Efficient Mobile Video Surveillance
|
| 著者名 |
|
◎Misa Nimura, Kenji Kanai, Jiro Katto,
|
| キーワード |
|
Q-learning, Path planning, Mobile sensing
|
| 抄録 |
|
In mobile sensing by vehicles, the vehicle should be navigated to hotspots efficiently while it travels to the destination, and it is necessary to maximize the amount of data transfer on the moving path in order to upload videos as fast as possible. However, throughputs of hotspots fluctuate because the number of accommodated users or sensors is dynamically changed. In this paper, we consider throughput fluctuation in the path planning task and search the optimal path by using Q-learning. By carrying out computer simulations, we evaluate the total data amount on the paths, then we can find the better paths.
|
| 本文pdf |
|
PDF download
|
