Immersive communication has been identified as a key application scenario in IMT-2030,encompassing emerging use cases such as immersive XR,the metaverse,and holographic communication.However,the high-resolution and high-frame-rate data transmission requirements of immersive applications pose significant challenges to existing communication network architectures.To address these challenges,this paper proposes an AI-native immersive communication network architecture based on semantic cognition and generation.This architecture aims to ensure user experience while significantly reducing transmission data volume and supporting personalized video rendering and transmission.Specifically,semantic cognition and inference mechanisms enable users to extract rich semantic knowledge from raw physical-world data.At the receiving end,generative AI models map semantic information into three-dimensional,multi-view video representations.An industrial digital twin scenario is used as a case study,building an intelligent robotic arm prototype platform to evaluate the performance of the proposed architecture.Experimental results demonstrate that the proposed approach significantly reduces data transmission load and end-to-end latency compared to traditional communication architectures.
关键词
6G/沉浸式通信/语义认知/生成式AI
Key words
6G/immersive communication/semantic cognition/generative AI
维普
万方数据