Multi-Modal Conditioned High-Resolution Transformer for Urban Electromagnetic Field Map Prediction Download PDF
For network planners needing fast EMF maps, this work provides a practical deep learning alternative to slow simulators, though the improvement is incremental over existing architectures.
The paper tackles computationally expensive physics-based electromagnetic field (EMF) map prediction for urban cellular network planning. Their multi-conditioned HRFormer model achieves a test MAE of 0.0461, improving 25.2% over a UNet baseline and 31.8% over HRFormer-only.
Predicting electromagnetic field (EMF) strength in urban environments is essential for cellular network planning but computationally expensive with physics-based simulators. We propose a multi-conditioned dense prediction framework that generates 500 500 EMF maps from building layout images and antenna configurations. Our architecture uses a High-Resolution Transformer (HRFormer) backbone with two complementary conditioning mechanisms: Feature-wise Linear Modulation (FiLM) injects scalar antenna parameters into all backbone stages, while cross-attention fuses 1-D radiation pattern tokens with spatial features at the deepest stage. We further introduce transmitter-relative spatial channels encoding distance, proximity, and bearing from the antenna, enabling coordinate-consistent test-time augmentation (TTA) that reduces test MAE by 6.3%. To address the prediction difficulty imbalance across EMF maps, we design a composite loss combining masked L1, multi-scale structural similarity (MS-SSIM), and a focal L1 term that upweights high-signal pixels, outperforming individual loss components in all metrics. Our best model achieves a test MAE of 0.0461, a 25.2% improvement over a plain UNet baseline and 31.8% over an HRFormer-only baseline.Do-