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 2020-08-25 10:50
Policy Gradient-based Deep Reinforcement Learning for Deadline-aware Data Transfer over Wide Area Networks
Masaki Notoya, Kohei Shiomoto (Tokyo City Univ.), Takashi Kurimoto (NII) PN2020-21
Abstract (in Japanese) (See Japanese page) 
(in English) Deadline-aware job scheduling problems have been attracting attention in the application domains of scientific workflows and commercial data-center wide area networks. The scheduling of deadline-aware jobs is not an easy task. Satisfying all of the demands before the corresponding deadlines requires optimally spreading traffic along future times. Existing methods formulate an optimization problem which in most cases is computationally expensive. The cumulative effect of running a sequence of optimization problems can be very poor since only limited knowledge of the future is available when each optimization problem is solved. In this paper, we leverage a deep reinforcement learning, Policy Gradient, to build a novel deadline-aware job scheduling method. Through the simulation, we obtain the following findings. The EDF-trained initial policy outperforms a random initial policy. The proposed method outperforms the existing methods including RR, FCFS, and EDF with wider gaps from the existing methods expand as the traffic load increases.
Keyword (in Japanese) (See Japanese page) 
(in English) Deadline / High Speed Network / Data Center / Job Scheduling / Deep Reinforcement Learning / Policy Gradient / /  
Reference Info. IEICE Tech. Rep., vol. 120, no. 137, PN2020-21, pp. 49-56, Aug. 2020.
Paper # PN2020-21 
Date of Issue 2020-08-17 (PN) 
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 PN2020-21

Conference Information
Committee PN  
Conference Date 2020-08-24 - 2020-08-25 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To PN 
Conference Code 2020-08-PN 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Policy Gradient-based Deep Reinforcement Learning for Deadline-aware Data Transfer over Wide Area Networks 
Sub Title (in English)  
Keyword(1) Deadline  
Keyword(2) High Speed Network  
Keyword(3) Data Center  
Keyword(4) Job Scheduling  
Keyword(5) Deep Reinforcement Learning  
Keyword(6) Policy Gradient  
1st Author's Name Masaki Notoya  
1st Author's Affiliation Tokyo City University (Tokyo City Univ.)
2nd Author's Name Kohei Shiomoto  
2nd Author's Affiliation Tokyo City University (Tokyo City Univ.)
3rd Author's Name Takashi Kurimoto  
3rd Author's Affiliation National Institute of Informatics (NII)
4th Author's Name  
4th Author's Affiliation ()
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 2020-08-25 10:50:00 
Presentation Time 25 minutes 
Registration for PN 
Paper # PN2020-21 
Volume (vol) vol.120 
Number (no) no.137 
Page pp.49-56 
Date of Issue 2020-08-17 (PN) 

[Return to Top Page]

[Return to IEICE Web Page]

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