ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Journal Article 무선 통신 물리 계층의 기계 학습 활용 동향
Cited - time in scopus Download 145 time Share share facebook twitter linkedin kakaostory
Authors
최윤호, 강현덕, 김도영, 이재호, 박윤옥
Issue Date
2018-04
Citation
전자통신동향분석, v.33, no.2, pp.39-47
ISSN
1225-6455
Publisher
한국전자통신연구원 (ETRI)
Language
Korean
Type
Journal Article
DOI
https://dx.doi.org/10.22648/ETRI.2018.J.330205
Abstract
The fundamental problem of communica-tion is that of transmitting a message from a source to a destination over a channel through the use of a transmitter and receiver. To derive a theoretically optimal solution, the transmitter and receiver can be divided into several processing blocks, with each component analyzed and optimized. The idea of machine learning (or deep learning) communications systems goes back to the original definition of the communi-cation problem, and optimizes the transmitter and receiver jointly. Alt-hough today’s systems have been opti-mized over the last decades, and it seems difficult to compete with their performance, deep learning based commu-nication is attractive owing to its simplicity and the fact that it can learn to communicate over any type of channel without the need for mathemati-cal modeling or analysis.
KSP Keywords
Communications system, Transmitter and receiver, deep learning(DL), machine Learning, need for, optimal solution
This work is distributed under the term of Korea Open Government License (KOGL)
(Type 4: : Type 1 + Commercial Use Prohibition+Change Prohibition)
Type 4: