Distributed processing method for deep learning in wireless sensor networks

In wireless sensor networks, distributed processing technology for deep learning that utilizes edge computing and mobile terminals has been attracting interest; however, with the expansion of the use of wireless sensor networks, the amount of data and the increase in server load are issues, especial...

Full description

Saved in:
Bibliographic Details
Published inIEICE COMMUNICATIONS EXPRESS Vol. 10; no. 8; pp. 505 - 510
Main Authors Umeda, Karin, Nishitsuji, Takashi, Asaka, Takuya, Miyoshi, Takumi
Format Journal Article
LanguageEnglish
Published The Institute of Electronics, Information and Communication Engineers 01.08.2021
Subjects
deep learning
distributed processing
wireless sensor network
Online AccessGet full text

Cover

More Information
Summary:In wireless sensor networks, distributed processing technology for deep learning that utilizes edge computing and mobile terminals has been attracting interest; however, with the expansion of the use of wireless sensor networks, the amount of data and the increase in server load are issues, especially for deep learning. Although the existing studies which improve the data processing speed and latency reduction of distributed processing, there is still a problem in which the amount of communication and the load on the server increase. In this paper, we propose load balancing algorithm for deep learning to reduce the communication volume and server load.
ISSN:2187-0136
2187-0136
DOI:10.1587/comex.2021ETL0029