Important Bit Prefix M-ary Quadrature Amplitude Modulation for Semantic Communications

TL;DR

This paper introduces IBP-MQAM, a novel modulation scheme for semantic communications, which improves performance by prioritizing important bits based on text semantics extracted via LDA.

Contribution

The paper proposes IBP-MQAM, a new modulation scheme tailored for semantic communication, with derived error rate expressions and demonstrated performance improvements.

Findings

IBP-MQAM outperforms traditional MQAM in semantic communication scenarios.

Extraction of text semantics with LDA enhances modulation performance.

Key system parameters significantly affect IBP-MQAM effectiveness.

Abstract

M-ary Quadrature Amplitude Modulation (MQAM) is a commonly used channel modulation technology in wireless communication systems. To achieve dedicated channel modulation for semantic communication (SemCom), we propose an Important-Bit-Prefixed MQAM (IBP-MQAM) scheme and derive its approximate expression of important symbol error rate (ISER) and unimportant symbol error rate (USER). By extracting and quantifying text semantics using Latent Dirichlet Allocation (LDA), we verify that IBP-MQAM achieves improved performance over MQAM in SemCom scenarios and further analyze the effects of key system parameters.