Ai-Based Predictive Channel Modeling Platform for 6G Multi-Band and Multi-Scenario Communications

Abstract

The future sixth-generation (6G) wireless communications will realize global coverage, full-spectrum utilization, and intelligent collaboration, laying the foundation for industrial internet of things (IIoT) applications. Channel modeling is the cornerstone for wireless communication system design and standardization, but accurate channel modeling for unknown frequency bands and IIoT scenarios remains a formidable challenge. To address the problem, this paper proposes and implements an artificial intelligence (AI)-based platform for intelligent predictive channel modeling. With a unified data processing kernel and a lightweight time-series prediction algorithm at its core, the platform integrates three mainstream algorithms: temporal convolutional network (TCN), long short-term memory (LSTM), and gated recurrent unit (GRU). The platform supports flexible configuration and completes standardized processing of heterogeneous data in multiple frequency bands and scenarios. Experimental results show that the platform can achieve highprecision prediction of channel characteristics within millisecondlevel latency in multiple scenarios. This work provides core technical support for intelligent 6G IIoT network deployment and accelerates the industrial application of 6G technologies.