Causal Semantic Transmission Architecture Research

Abstract: With the evolution of 6G communication toward semantic communication, the traditional bit transmission mode suffers excessive bandwidth redundancy and low transmission efficiency. To address the problems of insufficient causal constraints and severe semantic distortion in existing semantic transmission schemes, this paper proposes a semantic transmission architecture based on causal reasoning. The architecture constructs causal graphs to model the semantic correlation of services, extracts semantic features and compresses causal factors at the transmitter, and recovers semantics with causal constraints at the receiver to avoid the transmission of redundant bits. Simulation tests on a communication link verify that the proposed scheme effectively reduces bandwidth consumption and improves semantic restoration accuracy in complex scenarios, which can provide references for the design of lightweight 6G semantic communication systems.