GeoGS-CE: Learning Delay--Beam Channel Priors with 3D Gaussians for High-Mobility Scenarios

2026-05-15

AI Summary

This work addresses the challenge of accurate wideband channel estimation in high-mobility scenarios, where rapid temporal variations and sparse pilot signals hinder performance. The authors propose GeoGS-CE, a two-stage framework that first constructs an offline geometric channel model based on 3D Gaussian scattering and incorporates leakage-aware differentiable wireless rendering to jointly model non-line-of-sight and virtual line-of-sight paths. In the online phase, it predicts location-dependent delay-beam power spectra as covariance priors to drive linear MMSE reconstruction of full-bandwidth, full-array channels. By uniquely integrating 3D Gaussian representations with differentiable rendering, the method leverages environmental geometry to learn stable, geometry-rich priors, effectively mitigating the sensitivity to random phases inherent in conventional approaches. Simulations on the Guangzhou–Shenzhen high-speed railway demonstrate that the proposed geometric prior significantly outperforms pilot-only and non-geometric baselines in reconstruction accuracy.

Abstract

Wideband channel estimation (CE) in high-mobility scenarios remains challenging because channel responses vary rapidly, while practical systems can allocate only sparse pilots to accommodate dense users. Fortunately, many high-mobility environments, such as high-speed railways, exhibit scheduled trajectories, predictable velocities, and a limited number of dominant propagation paths. These properties induce a delay--beam power spectrum that is more stable than the instantaneous complex channel frequency response (CFR), less sensitive to the random phase coherence, and rich in geometric information. To exploit such environmental properties, we propose GeoGS-CE, a two-stage channel estimation framework for sparse-pilot high-mobility scenarios. In the offline stage, GeoGS-CE jointly models: 1) a scene-level 3D Gaussian representation that captures the non-line-of-sight (NLoS) geometric scattering support, and 2) a leakage-aware differentiable wireless rendering process that maps the NLoS Gaussians, together with an explicit virtual line-of-sight (LoS) component, to the measured delay--beam power spectrum, while accounting for practical OFDM delay and array leakage effects. In the online stage, the delay--beam power spectrum is predicted for each user location and used as a strong covariance prior, enabling accurate full-band and full-array CFR reconstruction and tracking through a linear MMSE estimator. Simulations based on channels generated from a segment of the Guangshen high-speed railway show that the proposed geometric prior substantially improves CFR reconstruction over pilot-only and non-geometric baselines.

Problem

• wideband channel estimation
• high-mobility scenarios
• sparse pilots
• delay-beam power spectrum
• channel frequency response

Innovation

• 3D Gaussian Splatting
• Geometric Channel Prior
• Delay–Beam Power Spectrum
• High-Mobility Channel Estimation
• Differentiable Wireless Rendering