Large-Language-Model-Enabled Text Semantic Communication Systems

Large language models (LLMs) have recently demonstrated state-of-the-art performance in various natural language processing (NLP) tasks, achieving near-human levels in multiple language understanding challenges and aligning closely with the core principles of semantic communication Inspired by LLMs’...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhenyi Wang, Li Zou, Shengyun Wei, Kai Li, Feifan Liao, Haibo Mi, Rongxuan Lai
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/13/7227
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849319678818648064
author Zhenyi Wang
Li Zou
Shengyun Wei
Kai Li
Feifan Liao
Haibo Mi
Rongxuan Lai
author_facet Zhenyi Wang
Li Zou
Shengyun Wei
Kai Li
Feifan Liao
Haibo Mi
Rongxuan Lai
author_sort Zhenyi Wang
collection DOAJ
description Large language models (LLMs) have recently demonstrated state-of-the-art performance in various natural language processing (NLP) tasks, achieving near-human levels in multiple language understanding challenges and aligning closely with the core principles of semantic communication Inspired by LLMs’ advancements in semantic processing, we propose LLM-SC, an innovative LLM-enabled semantic communication system framework which applies LLMs directly to the physical layer coding and decoding for the first time. By analyzing the relationship between the training process of LLMs and the optimization objectives of semantic communication, we propose training a semantic encoder through LLMs’ tokenizer training and establishing a semantic knowledge base via the LLMs’ unsupervised pre-training process. This knowledge base facilitates the creation of optimal decoder by providing the prior probability of the transmitted language sequence. Based on this, we derive the optimal decoding criteria for the receiver and introduce beam search algorithm to further reduce complexity. Furthermore, we assert that existing LLMs can be employed directly for LLM-SC without extra re-training or fine-tuning. Simulation results reveal that LLM-SC outperforms conventional DeepSC at signal-to-noise ratios (SNRs) exceeding 3 dB, as it enables error-free transmissions of semantic information under high SNRs while DeepSC fails to do so. In addition to semantic-level performance, LLM-SC demonstrates compatibility with technical-level performance, achieving approximately an 8 dB coding gain for a bit error ratio (BER) of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mn>10</mn><mrow><mo>−</mo><mn>3</mn></mrow></msup></semantics></math></inline-formula> without any channel coding while maintaining the same joint source–channel coding rate as traditional communication systems.
format Article
id doaj-art-313ee98e8cba4ed0a5dabfef0c166c9d
institution Kabale University
issn 2076-3417
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-313ee98e8cba4ed0a5dabfef0c166c9d2025-08-20T03:50:21ZengMDPI AGApplied Sciences2076-34172025-06-011513722710.3390/app15137227Large-Language-Model-Enabled Text Semantic Communication SystemsZhenyi Wang0Li Zou1Shengyun Wei2Kai Li3Feifan Liao4Haibo Mi5Rongxuan Lai6College of Information and Communication, National University of Defense Technology, Wuhan 430000, ChinaCollege of Information and Communication, National University of Defense Technology, Wuhan 430000, ChinaSchool of Electrical Engineering, Naval University of Engineering, Wuhan 430000, ChinaCollege of Information and Communication, National University of Defense Technology, Wuhan 430000, ChinaCollege of Information and Communication, National University of Defense Technology, Wuhan 430000, ChinaCollege of Information and Communication, National University of Defense Technology, Wuhan 430000, ChinaCollege of Information and Communication, National University of Defense Technology, Wuhan 430000, ChinaLarge language models (LLMs) have recently demonstrated state-of-the-art performance in various natural language processing (NLP) tasks, achieving near-human levels in multiple language understanding challenges and aligning closely with the core principles of semantic communication Inspired by LLMs’ advancements in semantic processing, we propose LLM-SC, an innovative LLM-enabled semantic communication system framework which applies LLMs directly to the physical layer coding and decoding for the first time. By analyzing the relationship between the training process of LLMs and the optimization objectives of semantic communication, we propose training a semantic encoder through LLMs’ tokenizer training and establishing a semantic knowledge base via the LLMs’ unsupervised pre-training process. This knowledge base facilitates the creation of optimal decoder by providing the prior probability of the transmitted language sequence. Based on this, we derive the optimal decoding criteria for the receiver and introduce beam search algorithm to further reduce complexity. Furthermore, we assert that existing LLMs can be employed directly for LLM-SC without extra re-training or fine-tuning. Simulation results reveal that LLM-SC outperforms conventional DeepSC at signal-to-noise ratios (SNRs) exceeding 3 dB, as it enables error-free transmissions of semantic information under high SNRs while DeepSC fails to do so. In addition to semantic-level performance, LLM-SC demonstrates compatibility with technical-level performance, achieving approximately an 8 dB coding gain for a bit error ratio (BER) of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msup><mn>10</mn><mrow><mo>−</mo><mn>3</mn></mrow></msup></semantics></math></inline-formula> without any channel coding while maintaining the same joint source–channel coding rate as traditional communication systems.https://www.mdpi.com/2076-3417/15/13/7227large language modeljoint source–channel codingjoint source–channel decodingsemantic communication
spellingShingle Zhenyi Wang
Li Zou
Shengyun Wei
Kai Li
Feifan Liao
Haibo Mi
Rongxuan Lai
Large-Language-Model-Enabled Text Semantic Communication Systems
Applied Sciences
large language model
joint source–channel coding
joint source–channel decoding
semantic communication
title Large-Language-Model-Enabled Text Semantic Communication Systems
title_full Large-Language-Model-Enabled Text Semantic Communication Systems
title_fullStr Large-Language-Model-Enabled Text Semantic Communication Systems
title_full_unstemmed Large-Language-Model-Enabled Text Semantic Communication Systems
title_short Large-Language-Model-Enabled Text Semantic Communication Systems
title_sort large language model enabled text semantic communication systems
topic large language model
joint source–channel coding
joint source–channel decoding
semantic communication
url https://www.mdpi.com/2076-3417/15/13/7227
work_keys_str_mv AT zhenyiwang largelanguagemodelenabledtextsemanticcommunicationsystems
AT lizou largelanguagemodelenabledtextsemanticcommunicationsystems
AT shengyunwei largelanguagemodelenabledtextsemanticcommunicationsystems
AT kaili largelanguagemodelenabledtextsemanticcommunicationsystems
AT feifanliao largelanguagemodelenabledtextsemanticcommunicationsystems
AT haibomi largelanguagemodelenabledtextsemanticcommunicationsystems
AT rongxuanlai largelanguagemodelenabledtextsemanticcommunicationsystems