Physics-Informed Deep Recurrent Back-Projection Network for Tunnel Propagation Modeling
Kunyu Wu 1 {}^{\textbf{1}}, Qiushi Zhao 1 {}^{\textbf{1}}, Jingyi Zhou 1 {}^{\textbf{1}}, Junqiao Wang 1 {}^{\textbf{1}}, Hao Qin 1, 2 {}^{\textbf{1, 2}}, Xinyue Zhang 2 {}^{\textbf{2}}, and Xingqi Zhang 3 {}^{\textbf{3}}, Senior Member, IEEE · 2026
Read Physics-Informed Deep Recurrent Back-Projection Network for Tunnel Propagation Modeling by Kunyu Wu 1 {}^{\textbf{1}}, Qiushi Zhao 1 {}^{\textbf{1}}, Jingyi Zhou 1 {}^{\textbf{1}}, Junqiao Wang 1 {}^{\textbf{1}}, Hao Qin 1, 2 {}^{\textbf{1, 2}}, Xinyue Zhang 2 {}^{\textbf{2}}, and Xingqi Zhang 3 {}^{\textbf{3}}, Senior Member, IEEE (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-02007.txt.
Description
Radio signals inside railway tunnels decide whether communication-based train control works, but predicting them at fine resolution with parabolic wave equation solvers is expensive, while coarse solvers smear out the details. Kunyu Wu and six co-authors build PRBPN, a physics-informed recurrent network that rebuilds fine signal-strength fields from coarse solver slices.
The network fuses multiple slices over time. It iterates projection and back-projection steps. Those steps enforce physical consistency. Any pre-upsampling stage is skipped. The design stays data-efficient. Across four tunnel shapes and four frequencies, it tracked fine-mesh references closely. Validation used a real French tunnel, the Massif Central. Robustness was confirmed there. Training had used only a few paired examples. The method promises high-fidelity tunnel coverage planning at a fraction of the usual computing cost.
Catalog record
| CREATOR | Kunyu Wu 1 {}^{\textbf{1}}, Qiushi Zhao 1 {}^{\textbf{1}}, Jingyi Zhou 1 {}^{\textbf{1}}, Junqiao Wang 1 {}^{\textbf{1}}, Hao Qin 1, 2 {}^{\textbf{1, 2}}, Xinyue Zhang 2 {}^{\textbf{2}}, and Xingqi Zhang 3 {}^{\textbf{3}}, Senior Member, IEEE |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-02007