Paper Abstract and Keywords |
Presentation |
2021-11-26 15:00
[Poster Presentation]
Frame-Capture-Based CSI Recomposition Pertaining to Firmware-Agnostic WiFi Sensing Ryosuke Hanahara, Sohei Itahara, Kota Yamashita (Kyoto Univ.), Yusuke Koda (Univ. Oulu), Akihito Taya (Kyoto Univ./Aoyama Gakuin Univ.), Takayuki Nishio (Kyoto Univ./Tokyo Tech.), Koji Yamamoto (Kyoto Univ.) SRW2021-48 SeMI2021-47 CNR2021-22 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
This paper presents an estimation method of channel state information (CSI) matrices using corresponding beamforming feedback matrices (BFMs) contained in WLAN frames. The BFM is a right singular matrix computed by singular value decomposition of the CSI matrix. The CSI matrix is processed and discarded at the physical layer, while the BFM is exchanged between the access point and station
and can be obtained by frame capturing. On the other hand, more sensing tasks are based on CSI matrices rather than BFM. To fill this gap, this method uses machine learning to reconstruct the CSI matrices from the BFMs. The simulation results show that the estimated CSI amplitude approximates the ground-truth,
and leveraging BFMs on multiple subcarriers enhances prediction accuracy. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
channel state information / beamforming feedback / singular value decomposition / wireless sensing / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 121, no. 266, SeMI2021-47, pp. 58-60, Nov. 2021. |
Paper # |
SeMI2021-47 |
Date of Issue |
2021-11-18 (SRW, SeMI, CNR) |
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 |
SRW2021-48 SeMI2021-47 CNR2021-22 |
|