LDPC陪集码的可区分性及其在语义通信中的应用  

On the Distinguishability of LDPC Coset Codes and Its Application in Semantic Communication

作　　者：王寅楚 马啸[1,2] WANG Yinchu;MA Xiao(School of Computer Science and Engineering,Guangzhou 510006,China;Guangdong Province Key Laboratory of Information Security Technology,Sun Yat-sen University,Guangzhou 510006,China)

机构地区：[1]中山大学计算机学院,广东广州510006 [2]广东省信息安全技术重点实验室,广东广州510006

出　　处：《移动通信》2023年第4期54-59,共6页Mobile Communications

基　　金：国家重点研发计划资助项目(2021YFA1000500)。

摘　　要：经典通信的目标是可靠传输信息符号,其很少利用语义信息。提出了一种在低时延的情况下,利用数据信息之间的语义关联性来获得经典译码性能增益的传输方案。基于低密度奇偶校验陪集码的可区分性,所提方案利用便车码来传输语义信息关联标识。若接收端判断信息存在语义关联,则根据预设的语义关联度进行迭代译码。仿真结果表明,语义信息能够为经典信道编码带来增益。The objective of the classical communication is to transmit reliably the information symbols,where the semantic information is rarely employed.A transmission scheme,under the constraint of low latency,is proposed to improve the conventional decoding by employing the semantic correlation between data.Based on the distinguishability of low-density parity-check(LDPC)coset codes,the proposed scheme uses free-ride coding to transmit the identifier of the semantic correlation.If the receiver detects a positive identifier,it will perform the iterative decoding algorithm based on a preset degree of semantic similarity.The simulation results show that the semantic information can bring gains to the classical channel coding.

关 键 词：经典通信 语义信息 低密度奇偶校验陪集码 便车码 语义关联 

分 类 号：TN911.22[电子电信—通信与信息系统]