What Is Semantic Communication?A View on Conveying Meaning in the Era of Machine Intelligence

作     者：Qiao Lan Dingzhu Wen Zezhong Zhang Qunsong Zeng Xu Chen Petar Popovski Kaibin Huang 

作者机构：Department of Electrical and Electronic EngineeringThe University of Hong KongHong Kong 999007China Department of Electronic SystemsAalborg UniversityAalborgDenmark 

出 版 物：《Journal of Communications and Information Networks》 (通信与信息网络学报（英文）)

年 卷 期：2021年第6卷第4期

页      面：336-371页

核心收录：

学科分类：080904[工学-电磁场与微波技术] 0810[工学-信息与通信工程] 0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 080402[工学-测试计量技术及仪器] 0804[工学-仪器科学与技术] 081001[工学-通信与信息系统] 

基　　金：a fellowship award from the Research Grants Council of Hong Kong Special Administrative Region,China(HKU RFS21227S04) Guangdong Basic and Applied Basic Research Foundation(2019B1515130003) Hong Kong Research Grants Council(17208319) Hong Kong Research Grants Council(17209917) the Innovation and Technology Fund(GHP/016/18GD) Shenzhen Science and Technology Program(JCYJ20200109141414409) 

主　　题：semantic communication artificial intelligence Internet-of-things 

摘      要：In the 1940s,Claude Shannon developed the information theory focusing on quantifying the maximum data rate that can be supported by a communication *** by this fundamental work,the main theme of wireless system design up until the fifth generation(5G)was the data rate *** Shannon’s theory,the semantic aspect and meaning of messages were treated as largely irrelevant to *** classic theory started to reveal its limitations in the modern era of machine intelligence,consisting of the synergy between Internet-of-things(IoT)and artificial intelligence(AI).By broadening the scope of the classic communication-theoretic framework,in this article,we present a view of semantic communication(SemCom)and conveying meaning through the communication *** address three communication modalities:human-to-human(H2H),human-to-machine(H2M),and machine-to-machine(M2M)*** latter two represent the paradigm shift in communication and computing,and define the main theme of this article.H2M SemCom refers to semantic techniques for conveying meanings understandable not only by humans but also by machines so that they can have interaction and“dialogue.On the other hand,M2M SemCom refers to effective techniques for efficiently connecting multiple machines such that they can effectively execute a specific computation task in a wireless *** first part of this article focuses on introducing the SemCom principles including encoding,layered system architecture,and two design approaches:1)layer-coupling design;and 2)end-to-end design using a neural *** second part focuses on the discussion of specific techniques for different application areas of H2M SemCom[including human and AI symbiosis,recommendation,human sensing and care,and virtual reality(VR)/augmented reality(AR)]and M2M SemCom(including distributed learning,split inference,distributed consensus,and machine-vision cameras).Finally,we discuss the approach for designing S