A new research paper published in Engineering presents wireless environmental information theory (WEIT) as a novel theoretical foundation for 6G environment intelligence communication (EIC), offering a proactive and online system design paradigm to address the limitations of traditional statistical channel models in mobile communications. Authored by a research team from Beijing University of Posts and Telecommunications and China Mobile Research Institute, the study establishes WEIT, validates the EIC architecture aided by wireless environmental information (WEI), and discusses key challenges for its practical 6G application, marking a meaningful exploration of integrating communication, sensing and artificial intelligence (AI) in next-generation wireless networks.

The statistical channel modeling paradigm has guided 1G to 5G communication system design, yet it relies on offline measurements in limited typical environments, leading to passive system adaptation and suboptimal performance in dynamic real-world scenarios. As 6G targets ubiquitous connectivity and higher capacity with scenarios like integrated sensing and communication (ISAC), the research team proposes EIC based on WEIT to realize real-time intelligent interaction between communication systems and the physical environment. The EIC framework operates through three core steps: acquiring WEI via multimodal sensing techniques, predicting channel fading using AI with WEI and channel data, and enabling the system to autonomously determine optimal air-interface transmission strategies based on real-time channel predictions.

To underpin EIC, the study defines WEI as the physical properties of environmental objects and scatterers that influence wireless channel characteristics, classifying it into static, dynamic and random categories, and identifying its key properties of homogeneity, consistency and correlation. It also quantifies WEI with wireless environmental entropy, establishing the relationship between environmental uncertainty and channel determinacy from the perspective of system capacity. The proposed EIC-WEI architecture features a closed-loop process including multimodal sensing and environment reconstruction, knowledge mapping, AI-based channel fading prediction, proactive decision-making and optimal transmission strategy implementation, with a detailed five-step WEI flow from raw environmental data collection to wireless environmental knowledge (WEK) construction.

The research validates EIC-WEI across four air-interface tasks: cell coverage, channel state information (CSI) prediction, optimal beam selection and air interface resource management. Simulation results show EIC-WEI outperforms traditional statistical models in reducing overhead and optimizing performance, such as decreasing the normalized mean square error (NMSE) of small-scale channel parameter prediction by approximately 59.8% and improving the top 3 beam prediction accuracy by 29%. It also achieves fair multi-user resource allocation without sacrificing total system throughput.

The study further outlines key challenges for EIC-WEI’s practical application, including improving the accuracy of WEI acquisition for stable system performance, reducing the complexity of environment interaction to support real-time 6G tasks, and enhancing environmental generalization to achieve ubiquitous intelligence. Potential solutions involve multimodal WEI fusion, model compression techniques, WEK construction and the development of channel large models with federated learning, laying a foundation for future 6G environment intelligence communication research and implementation.

The paper “Wireless Environmental Information Theory: A New Paradigm Toward 6G Online and Proactive Environment Intelligence Communication,” is authored by Jianhua Zhang, Li Yu, Shaoyi Liu, Yichen Cai, Yuxiang Zhang, Hongbo Xing, Tao Jiang. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.07.028. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.