3D Dynamic Radio Map Prediction Using Vision Transformers for Low-Altitude Wireless Networks

Low-altitude wireless networks (LAWN) are expanding rapidly with the deployment of unmanned aerial vehicles (UAVs) for diverse applications. However, ensuring reliable connectivity for a large number of cellular-connected UAVs is challenging due to their high 3D mobility and time-varying traffic demands. These dynamics force base stations (BSs) to adapt their transmit power in real-time, creating a highly non-stationary 3D radio environment. While radio maps (RMs) are effective for spatial characterization, existing static models fail to capture the real-time power evolution and spatio-temporal dependencies inherent in LAWNs. To overcome this limitation, we propose a 3D dynamic radio map (3D-DRM) framework that learns and predicts the spatio-temporal evolution of received power. Specially, a Vision Transformer (ViT) encoder extracts high-dimensional spatial representations from 3D RMs, while a Transformer-based module models sequential dependencies to predict future power distributions. Experiments unveil that 3D-DRM accurately captures fast-varying power dynamics and substantially outperforms baseline models in both RM reconstruction and short-term prediction.