Differentiable Ray Tracing with Gaussians for Unified Radio Propagation Simulation and View Synthesis

2026-05-08

AI Summary

This work addresses the limitations of existing neural scene representations, which prioritize visual appearance and struggle to support deterministic multipath tracing required for radio frequency (RF) propagation, as well as conventional RF simulation methods that rely on manually constructed meshes and lack a unified representation with visual reconstruction. To bridge this gap, the authors propose embedding 3D Gaussian primitives into a hardware-accelerated ray tracing framework, establishing a differentiable, unified model that simultaneously achieves high-quality novel view synthesis and physically plausible RF channel impulse response simulation. Their approach enables, for the first time, direct extraction of multipath RF trajectories from purely vision-driven neural scenes without additional geometric modeling, thereby demonstrating the feasibility and potential of neural representations for RF digital twins.

Abstract

Explicit neural representations such as 3D Gaussian Splatting (3DGS) enable high-fidelity and real-time novel view synthesis, yet optimize for alpha-composited optical appearance rather than ray-intersectable geometry. In contrast, radio-frequency (RF) digital twins require deterministic multi-bounce paths, where the geometry dictates trajectories and their associated attenuation and delay. We introduce a framework enabling differentiable RF propagation simulation directly within visually reconstructed neural scenes, allowing point-to-point path computation between arbitrary 3D locations while preserving high-quality visual rendering. Unlike conventional RF simulation pipelines that rely on manually constructed meshes, we embed Gaussian primitives into a hardware-accelerated ray tracing structure as the underlying spatial representation. By extracting physically meaningful channel impulse responses from visual-only reconstructions, we provide cross-modal evidence that neural reconstructions can serve as unified spatial representations for both electromagnetic propagation simulation and photorealistic view synthesis.

Problem

• Differentiable Ray Tracing
• 3D Gaussian Splatting
• Radio Propagation Simulation
• Neural Scene Representation
• View Synthesis

Innovation

• Differentiable Ray Tracing
• 3D Gaussian Splatting
• Radio Propagation Simulation
• Neural Scene Representation
• Cross-modal Reconstruction