arrow
返回

Semantics Alignment via Split Learning for Resilient Multi-User Semantic Communication

delete2024-10-01
delete0
delete
OA
AI
J
Jinhyuk Choi
J
Jihong Park *
S
Seung‐Woo Ko *
J
Jinho Choi
M
Mehdi Bennis
S
Seong‐Lyun Kim *
DOI:10.1109/TVT.2024.3410380delete
delete原文链接
delete分享
delete收藏
摘要

摘要

En 中文
Recent studies on semantic communication commonly rely on neural network (NN) based transceivers such as deep joint source and channel coding (DeepJSCC). Unlike traditional transceivers, these neural transceivers are trainable using actual source data and channels, enabling them to extract and communicate semantics. On the flip side, each neural transceiver is inherently biased towards specific source data and channels, making different transceivers difficult to understand intended semantics, particularly upon their initial encounter. To align semantics over multiple neural transceivers, we propose a distributed learning based solution, which leverages split learning (SL) and partial NN fine-tuning techniques. In this method, referred to as SL with layer freezing (SLF), each encoder downloads a misaligned decoder, and locally fine-tunes a fraction of these encoder-decoder NN layers. By adjusting this fraction, SLF controls computing and communication costs. Simulation results confirm the effectiveness of SLF in aligning semantics under different source data and channel dissimilarities, in terms of classification accuracy, reconstruction errors, and recovery time for comprehending intended semantics from misalignment.
Keyword:
Semantics
Transceivers
Task analysis
Artificial neural networks
Decoding
Training
Error probability
DeepJSCC
fine-tuning
neural transceiver
semantic communication
split learning

期刊

IEEE Transactions on Vehicular Technology 封面图
IEEE Transactions on Vehicular Technology
IF:
7.1
论文数:
1.8W
被引数:
6.6W

机构

U
University of Oulu
学者数:
1.5W
论文数: 1.3W
被引数: 1.6W
I
Inha University
学者数:
1.1W
论文数: 1.1W
被引数: 1.1W
Y
Yonsei University
学者数:
4.8W
论文数: 4.6W
被引数: 5.2W
D
Deakin University
学者数:
2.0W
论文数: 2.1W
被引数: 2.8W
学者 查看更多机构
引用论文

引用论文

Communication-Efficient and Distributed Learning Over Wireless Networks: Principles and Applications
err2021-05-01
err117
errOAAI
errPark, Jihong; Samarakoon, Sumudu; Elgabli, Anis; Kim, Joongheon; Bennis, Mehdi; Kim, Seong-Lyun; Debbah, Merouane
err分享
err收藏
err分享
err收藏
Semantics-Native Communication via Contextual Reasoning
err2023-06-01
err10
errOAAI
errSeo, Hyowoon; Park, Jihong; Bennis, Mehdi; Debbah, Merouane
err分享
err收藏
没有更多内容