ARTFEED — Contemporary Art Intelligence

GeoGS-CE: 3D Gaussian Channel Estimation for High-Mobility Wireless

other · 2026-05-18

A research paper on arXiv proposes GeoGS-CE, a two-stage channel estimation framework for high-mobility wireless scenarios like high-speed railways. The method leverages delay-beam power spectrum stability and geometric information from scheduled trajectories. In an offline stage, it jointly models a scene-level 3D Gaussian representation for non-line-of-sight scattering and a leakage-aware differential model. The approach aims to improve wideband channel estimation with sparse pilots, addressing rapid channel variations in dense user environments.

Key facts

  • GeoGS-CE is a two-stage channel estimation framework
  • Targets high-mobility scenarios such as high-speed railways
  • Uses delay-beam power spectrum that is more stable than instantaneous CFR
  • Offline stage models 3D Gaussian representation for NLoS scattering
  • Also includes a leakage-aware differential model
  • Aims to improve wideband channel estimation with sparse pilots
  • Published on arXiv with ID 2605.16094
  • Addresses rapid channel variations in dense user environments

Entities

Institutions

  • arXiv

Sources