Saved in:
Bibliographic Details
Main Authors: Yang, Kang, Chen, Yuning, Du, Wan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.05708
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918453414199296
author Yang, Kang
Chen, Yuning
Du, Wan
author_facet Yang, Kang
Chen, Yuning
Du, Wan
contents We present GRaF, Generalizable Radio-Frequency (RF) Radiance Fields, a framework that models RF signal propagation to synthesize spatial spectra at arbitrary transmitter or receiver locations, where each spectrum measures signal power across all surrounding directions at the receiver. Unlike state-of-the-art methods that adapt vanilla Neural Radiance Fields (NeRF) to the RF domain with scene-specific training, GRaF generalizes across scenes to synthesize spectra. To enable this, we prove an interpolation theory in the RF domain: the spatial spectrum from a transmitter can be approximated using spectra from geographically proximate transmitters. Building on this theory, GRaF comprises two components: (i) a geometry-aware Transformer encoder that captures spatial correlations from neighboring transmitters to learn a scene-independent latent RF radiance field, and (ii) a neural ray tracing algorithm that estimates spectrum reception at the receiver. Experimental results demonstrate that GRaF outperforms existing methods on single-scene benchmarks and achieves state-of-the-art performance on unseen scene layouts.
format Preprint
id arxiv_https___arxiv_org_abs_2502_05708
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Generalizable Radio-Frequency Radiance Fields for Spatial Spectrum Synthesis
Yang, Kang
Chen, Yuning
Du, Wan
Networking and Internet Architecture
Machine Learning
We present GRaF, Generalizable Radio-Frequency (RF) Radiance Fields, a framework that models RF signal propagation to synthesize spatial spectra at arbitrary transmitter or receiver locations, where each spectrum measures signal power across all surrounding directions at the receiver. Unlike state-of-the-art methods that adapt vanilla Neural Radiance Fields (NeRF) to the RF domain with scene-specific training, GRaF generalizes across scenes to synthesize spectra. To enable this, we prove an interpolation theory in the RF domain: the spatial spectrum from a transmitter can be approximated using spectra from geographically proximate transmitters. Building on this theory, GRaF comprises two components: (i) a geometry-aware Transformer encoder that captures spatial correlations from neighboring transmitters to learn a scene-independent latent RF radiance field, and (ii) a neural ray tracing algorithm that estimates spectrum reception at the receiver. Experimental results demonstrate that GRaF outperforms existing methods on single-scene benchmarks and achieves state-of-the-art performance on unseen scene layouts.
title Generalizable Radio-Frequency Radiance Fields for Spatial Spectrum Synthesis
topic Networking and Internet Architecture
Machine Learning
url https://arxiv.org/abs/2502.05708