PropSplat: Map-Free RF Field Reconstruction via 3D Gaussian Propagation Splatting

2026-05-08

AI Summary

This work addresses the high deployment cost and limited adaptability of traditional radio frequency (RF) propagation modeling, which typically relies on high-fidelity 3D maps or dense measurements and struggles in geospatially data-scarce environments. The paper introduces, for the first time, 3D Gaussian Splatting into map-free RF modeling, proposing an end-to-end learnable framework that reconstructs the propagation field using sparsely sampled RF measurements alone—without requiring prior knowledge of buildings or terrain. The approach employs anisotropic 3D Gaussian primitives initialized along transmitter–receiver paths and a learnable path-loss exponent. Evaluated on an outdoor sub-6 GHz dataset, the method achieves an RMSE of 5.38 dB, substantially outperforming existing techniques; in indoor BLE-based localization, it attains a mean error of merely 0.19 meters, improving upon NeRF² by nearly an order of magnitude.

Abstract

Building a site-specific propagation model typically requires either ray-tracing over detailed 3D maps or dense measurement campaigns. Both approaches are expensive and often infeasible for rapid deployments where geographic data is unavailable or outdated. We present PropSplat, a map-free propagation modeling method that reconstructs radio frequency (RF) fields using 3D anisotropic Gaussian primitives. Each Gaussian encodes a scalar path loss offset relative to an explicit baseline path loss model with a learnable path loss exponent. Gaussians are initialized along observed transmitter--receiver paths and optimized end-to-end to learn the propagation environment without external information like floor plans, terrain databases, or clutter data. We evaluate PropSplat against wireless radiance field methods NeRF$^2$, GSRF, and WRF-GS+ on two real-world datasets. On large-scale outdoor drive-tests spanning multiple topographical regions at six sub-6 GHz frequencies, PropSplat achieves 5.38 dB RMSE when training measurements are spaced 300m apart and outperforms WRF-GS+ (5.87 dB), GSRF (7.46 dB), and NeRF$^2$ (14.76 dB). On indoor Bluetooth Low Energy measurements, PropSplat achieves 0.19m mean localization error, an order of magnitude better than NeRF$^2$ (1.84m), while achieving near-identical received signal strength prediction accuracy. These results show that accurate site-specific propagation reconstruction is achievable from sparse RF-native measurements. The need for geographic data as a prerequisite for scalable RF environment modeling is reduced.

Problem

• RF field reconstruction
• map-free propagation modeling
• site-specific propagation
• wireless signal prediction
• sparse RF measurements

Innovation

• map-free propagation modeling
• 3D Gaussian splatting
• RF field reconstruction
• anisotropic Gaussians
• path loss modeling