Paper Abstract and Keywords |
Presentation |
2024-01-19 11:05
5G throughput prediction for 28GHz cell area using surrounding spatial information Hisashi Nagata, Riichi Kudo, kahoko takahashi, Fujita Takafumi (NIPPON TELEGRAPH AND TELEPHONE CORPORATION), Yuya Aoki, Morihiro Yoshifumi (NTT DOCOMO, INC.) SeMI2023-67 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
Every thing is connected to the network, and the amount of mobile traffic is increasing year by year. Therefore, the use of millimeter-wave is expected. However, millimeter-wave has problems with communication stability, and the link quality may deteriorate significantly due to the influence of objects surrounding the terminal. In order to stably utilize millimeter-wave bands, we believe that it is necessary to predict future LQ predictions and adaptively control wireless communication. 5G throughput prediction method has been proposed that uses the position, speed, and direction of the terminal and its surrounding objects as spatial information, but it is necessary to use different prediction models depending on the number of surrounding objects. In this paper, we propose a 5G prediction model that responds to changes in the number of surrounding objects. For evaluation, we constructed an indoor experimental environment in which a pedestrian has terminal connected to 5G network, and there are other pedestrians. We prepared an autonomous robot as a substitute for pedestrians and efficiently collected the data needed for learning. We prepared three patterns of up to three pedestrians and compared the prediction errors between a method that creates a prediction model for each and a proposed method that a common prediction model for these three patterns. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
5G / millimeter-wave / spatial information / communication link quality prediction / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 123, no. 345, SeMI2023-67, pp. 94-99, Jan. 2024. |
Paper # |
SeMI2023-67 |
Date of Issue |
2024-01-11 (SeMI) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
SeMI2023-67 |
|