基于语义通信的多源数据融合与重建

Multi-source Data Fusion and Reconstruction Based on Semantic Communication

6G网络需处理多模态数据并满足高实时性要求。随着数据流量快速增长,传统通信模式面临诸多挑战。语义通信通过在语义层面传递有意义的信息,有效减少带宽需求并提升通信效率,展现了巨大的应用前景。然而,现有的语义通信方案主要针对单一数据源,难以满足6G网络对多样化数据的传输需求。为此,设计了一种新型的多源数据融合语义通信系统。该系统通过在发射端融合不同数据源的语义特征,并在接收端重构这些特征,有效消除了数据间的冗余。实验结果表明,与传统的单一数据源通信方法相比,所提数据融合系统在相同带宽下具有更优性能,且在较低带宽条件下表现更加突出。

6G networks need to handle multi-modal data and meet the demand for high real-time requirements.With the rapid growth of data traffic,traditional communication paradigms face significant challenges.Semantic communication,which transmits meaningful information in the semantic domain,effectively reduces the required bandwidth and improves communication efficiency,making it a highly promising solution.However,existing semantic communication methods are primarily designed for a single data source,making them difficult to meet the diverse data transmission requirements of 6G networks.To address this issue,we propose a novel multi-source data fusion semantic communication system.This system fuses the semantic features from different sources at the transmitter and reconstructs these features at the receiver,effectively eliminating the redundancy between different data.Experimental results demonstrate that compared with traditional single-source communication methods,the proposed data fusion system performs better under the same bandwidth and excels even more in the scenario with lower bandwidth.