Research on High-Dynamic Wireless Environment Prediction and Reconstruction for Complex 6G Application Scenarios

The sixth generation (6G) mobile communication system imposes higher requirements on data rate, reliability, latency, and connection density. However, the convergence of multiple frequency bands, diverse scenarios, and emerging technologies makes wireless channels highly dynamic, while conventional offline statistical modeling methods can no longer meet the demands for real-time performance and generalization capability. To address this issue, this paper develops a new paradigm for wireless environment prediction and reconstruction based on “environment sensing → channel prediction → optimization decision-making”, and proposes an adaptive closed-loop theoretical framework from the physical environment to communication strategies. Based on this theoretical framework, a prototype experimental system for wireless environment prediction and reconstruction in complex 6G application scenarios is developed. The system integrates functional modules, including multimodal environment-channel joint acquisition, environment-aware digital twin channel prediction, and reconfigurable intelligent surface (RIS) intelligent control, enabling spatiotemporal alignment between heterogeneous sensing data and channel data, three-dimensional reconstruction of dynamic environments, sub-second channel prediction, and prediction-driven beam control. In a dynamic factory scenario, the effectiveness of the digital-twin-enabled RIS intelligent control method is further validated, achieving an average received signal power gain of approximately 10 dB. The results demonstrate that the proposed paradigm can support channel prediction, coverage enhancement, and adaptive optimization in high-dynamic wireless environments, providing new insights and key support for building intelligent, efficient, and evolvable 6G communication networks.