IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2022-06-17 10:35
A Study on Data-Driven Fine-Tuning for ICA-Aided Blind Signal Separation
Taisuke Nogami, Shinsuke Ibi (Doshisha Univ.), Takumi Takahashi (Osaka Univ.), Hisato Iwai (Doshisha Univ.) RCS2022-61
Abstract (in Japanese) (See Japanese page) 
(in English) Internet of Things (IoT) communications, which will plays an important role in the next-generation wireless communications infrastructure, will support massive connection requests, while packet sizes tend to be small.
MIMO (Multi-Input Multi-Output) transmission, which enables spatial multiplexing of signals at the same time and frequency, is a promising method to enable the large-scale connections. Generally, to separate spatially multiplexed signals, channel state information (CSI) is estimated by using pilot signals, and spatial filtering is applied based on the estimated CSI. However, it is undesirable from the viewpoint of transmission efficiency to add long pilot sequences to each small data packets. In such a situation, independent component analysis (ICA), a type of blind signal separation (BSS) technology, can be applied; however, ICA requires evaluations of non-Gaussianity and a relatively long data sequence for this estimation, so high detection accuracy cannot be expected even if ICA is applied directly to small packet size.
In this paper, we apply Deep Unfolding (DU), a type of deep learning, and improve the signal separation performance of ICA with relatively short data series length.
Keyword (in Japanese) (See Japanese page) 
(in English) ICA / deep learning / deep unfolding / data-driven tuning / / / /  
Reference Info. IEICE Tech. Rep., vol. 122, no. 73, RCS2022-61, pp. 218-223, June 2022.
Paper # RCS2022-61 
Date of Issue 2022-06-08 (RCS) 
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 RCS2022-61

Conference Information
Committee RCS  
Conference Date 2022-06-15 - 2022-06-17 
Place (in Japanese) (See Japanese page) 
Place (in English) University of the Ryukyus, Senbaru Campus and online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) First Presentation in IEICE Technical Committee, Resource Control, Scheduling, Wireless Communications, etc. 
Paper Information
Registration To RCS 
Conference Code 2022-06-RCS 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) A Study on Data-Driven Fine-Tuning for ICA-Aided Blind Signal Separation 
Sub Title (in English)  
Keyword(1) ICA  
Keyword(2) deep learning  
Keyword(3) deep unfolding  
Keyword(4) data-driven tuning  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Taisuke Nogami  
1st Author's Affiliation Doshisha University (Doshisha Univ.)
2nd Author's Name Shinsuke Ibi  
2nd Author's Affiliation Doshisha University (Doshisha Univ.)
3rd Author's Name Takumi Takahashi  
3rd Author's Affiliation Osaka University (Osaka Univ.)
4th Author's Name Hisato Iwai  
4th Author's Affiliation Doshisha University (Doshisha Univ.)
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Speaker Author-1 
Date Time 2022-06-17 10:35:00 
Presentation Time 10 minutes 
Registration for RCS 
Paper # RCS2022-61 
Volume (vol) vol.122 
Number (no) no.73 
Page pp.218-223 
#Pages
Date of Issue 2022-06-08 (RCS) 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan