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 2022-01-20 15:10
[Short Paper] A Study of Joint Control of Machine Learning Model and Wireless LAN Parameters in Split inference by Reinforcement Learning
Kojin Yorita (Tokyo Tech.), Sohei Itahara (Kyoto Univ.), Takayuki Nishio (Tokyo Tech.), Daiki Yoda, Toshihisa Nabetani (Toshiba) SeMI2021-66
Abstract (in Japanese) (See Japanese page) 
(in English) Distributed inference (DI) enables machine learning (ML) inference with a deep neural network on resource-constrained devices. However, lossy wireless networks can become a bottleneck, thereby increasing latency in DI. This paper studies a joint control of the wireless communication parameters (e.g., transmission rate, retransmission limit) and ML model architecture to reduce latency while maintaining inference accuracy. The proposed method focuses on the packet-loss tolerance of ML inference and uses unreliable (i.e., high packet-loss rate) but low-latency communication protocol. To achieve a well-balanced trade-off between accuracy and latency, the proposed method jointly controls wireless communication parameters affecting the reliability and latency and ML model architecture affecting accuracy and packet-loss reliance, based on multi-armed bandit (MAB) algorithm, namely upper confidence bound (UCB) algorithm. The results of ns3-ai-based computer simulations show that the proposed method reduces communication latency while maintaining inference accuracy.
Keyword (in Japanese) (See Japanese page) 
(in English) Distributed inference / Joint control / Machine learning / Wireless LAN / Reinforcement Learning / / /  
Reference Info. IEICE Tech. Rep., vol. 121, no. 333, SeMI2021-66, pp. 51-54, Jan. 2022.
Paper # SeMI2021-66 
Date of Issue 2022-01-13 (SeMI) 
ISSN Online edition: ISSN 2432-6380
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 SeMI2021-66

Conference Information
Committee SeMI  
Conference Date 2022-01-20 - 2022-01-21 
Place (in Japanese) (See Japanese page) 
Place (in English)  
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To SeMI 
Conference Code 2022-01-SeMI 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) A Study of Joint Control of Machine Learning Model and Wireless LAN Parameters in Split inference by Reinforcement Learning 
Sub Title (in English)  
Keyword(1) Distributed inference  
Keyword(2) Joint control  
Keyword(3) Machine learning  
Keyword(4) Wireless LAN  
Keyword(5) Reinforcement Learning  
1st Author's Name Kojin Yorita  
1st Author's Affiliation Tokyo Institute of Technology (Tokyo Tech.)
2nd Author's Name Sohei Itahara  
2nd Author's Affiliation Kyoto University (Kyoto Univ.)
3rd Author's Name Takayuki Nishio  
3rd Author's Affiliation Tokyo Institute of Technology (Tokyo Tech.)
4th Author's Name Daiki Yoda  
4th Author's Affiliation Toshiba (Toshiba)
5th Author's Name Toshihisa Nabetani  
5th Author's Affiliation Toshiba (Toshiba)
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 2022-01-20 15:10:00 
Presentation Time 10 minutes 
Registration for SeMI 
Paper # SeMI2021-66 
Volume (vol) vol.121 
Number (no) no.333 
Page pp.51-54 
Date of Issue 2022-01-13 (SeMI) 

[Return to Top Page]

[Return to IEICE Web Page]

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