PSC: A Packet Semantic Communication Framework for Lossy Transmission With Packet Loss Channel

IF 4.4 3区计算机科学 Q2 TELECOMMUNICATIONS

IEEE Communications Letters Pub Date : 2025-07-10 DOI:10.1109/LCOMM.2025.3587319

Wenyu Zhang;Zhiwei Xie;Sherali Zeadally;Kaiyuan Bai;Hua Shao;Haijun Zhang;Victor C. M. Leung

引用次数: 0

Abstract

Currently, semantic communication is predominantly confined to wireless communication scenarios, but not the general scenarios that transmit data packets. We propose packet semantic communication (PSC), a new semantic communication framework which achieves reliable transmissions in networks with packet loss. PSC works at the application layer, and it encodes the source data as binary data bits by using deep joint source-channel coding (DeepJSCC) and entropy coding, then transmits data as packets, and the decoder at the receiver performs the decoding process even when the packets are partially lost. We implement the proposed PSC for image transmission with a packet loss channel and compare its performance with the conventional better portable graphics (BPG) codec and learned image compression (LIC) model. The results obtained show that BPG and LIC are highly vulnerable to packet loss, but the proposed PSC can achieve strong robustness against packet loss, even when the packet loss rate (PLR) is as high as 80%.

查看原文本刊更多论文

具有丢包信道的有损传输的包语义通信框架

目前，语义通信主要局限于无线通信场景，而不是传输数据包的一般场景。本文提出了一种新的语义通信框架——分组语义通信（PSC），它可以在存在丢包的网络中实现可靠的传输。PSC工作在应用层，利用深度联合源信道编码（DeepJSCC）和熵编码将源数据编码为二进制数据位，然后以数据包的形式传输，即使数据包部分丢失，接收端的解码器也会执行解码过程。我们实现了带有丢包通道的PSC图像传输，并将其性能与传统的更好的便携式图形（BPG）编解码器和学习图像压缩（LIC）模型进行了比较。结果表明，BPG和LIC极易受到丢包的影响，而PSC即使在丢包率高达80%的情况下也能获得较强的抗丢包鲁棒性。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

IEEE Communications Letters 工程技术-电信学

CiteScore

8.10

自引率

7.30%

发文量

590

审稿时长

2.8 months

期刊介绍： The IEEE Communications Letters publishes short papers in a rapid publication cycle on advances in the state-of-the-art of communication over different media and channels including wire, underground, waveguide, optical fiber, and storage channels. Both theoretical contributions (including new techniques, concepts, and analyses) and practical contributions (including system experiments and prototypes, and new applications) are encouraged. This journal focuses on the physical layer and the link layer of communication systems.