pith. sign in

arxiv: 2605.22361 · v1 · pith:P7M5KPEZnew · submitted 2026-05-21 · 📡 eess.SP

Propagation-Consistent Wireless Environment Digital Twin Construction Under Sparse Measurements

Junjie Ai , Shurui Xu , Yanqing Ren , Zhuoyu Liu , Weicong Chen , Wankai Tang , Xiao Li , Chao-Kai Wen
show 1 more author
Shi Jin
This is my paper

Pith reviewed 2026-05-22 03:55 UTC · model grok-4.3

classification 📡 eess.SP
keywords wireless digital twinsparse measurementselectromagnetic calibrationray tracingchannel predictionBayesian channel mappropagation modelingscene reconstruction
0
0 comments X

The pith

Calibrating a scene-level electromagnetic property field on a geometric digital twin using sparse channel measurements produces a propagation-consistent wireless environment representation.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper aims to establish that a wireless environment digital twin can be constructed from sparse real measurements by evolving a pre-built geometric model through targeted calibration of its electromagnetic properties. It converts limited position-labeled channel data into dense probabilistic supervision via a geometry-prior Bayesian channel map, then optimizes the properties by embedding them in differentiable ray-tracing computations. A sympathetic reader would care because this yields channel predictions that hold for new device placements and supports planning plus data generation without exhaustive measurement campaigns.

Core claim

Instead of directly fitting link-specific channel responses, the method constructs a geometry-prior Bayesian channel map from sparse CSI to supply dense probabilistic targets with uncertainty, then embeds a learnable scene-level electromagnetic property field into differentiable ray tracing to evolve the geometric DT into a propagation-consistent WEDT.

What carries the argument

The learnable scene-level electromagnetic property field embedded in differentiable ray tracing and guided by a geometry-prior Bayesian channel map from sparse CSI.

If this is right

  • The calibrated twin delivers accurate channel prediction from sparse measurements.
  • Predictions generalize to transceiver topologies not encountered during calibration.
  • Performance holds across varying sampling densities and conditions.
  • The EM property field supports material-related environment sensing.
  • Higher-quality synthetic data can be generated for wireless AI applications.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach may allow incremental updates to the twin as fresh sparse measurements arrive without rebuilding the geometry.
  • It could reduce reliance on dense site surveys for large-scale wireless network planning.
  • Calibrated twins might supply more physically consistent training data for machine-learning models in communications.
  • Similar calibration could extend to dynamic scenes where objects or materials change over time.

Load-bearing premise

The initial geometric digital twin is accurate enough that calibrating only the scene-level electromagnetic property field produces propagation-consistent results.

What would settle it

Direct measurements of channel responses at unseen transceiver locations that deviate substantially from the predictions of the calibrated WEDT would show the claim of accurate and generalizable prediction is incorrect.

Figures

Figures reproduced from arXiv: 2605.22361 by Chao-Kai Wen, Junjie Ai, Shi Jin, Shurui Xu, Wankai Tang, Weicong Chen, Xiao Li, Yanqing Ren, Zhuoyu Liu.

Figure 1
Figure 1. Figure 1: Conceptual workflow for constructing a WEDT from a real-world environment. Multimodal sensing data collected from the real-world environment are used to reconstruct the geometric DT, while sparse position-labeled CSI data provide real propagation observations. The geometric DT is further evolved into the WEDT through a two-stage process. The WEDT has an EM property representation consistent with real-world… view at source ↗
Figure 2
Figure 2. Figure 2: Path propagation and interaction. The wireless signal is radiated by the Tx, undergoes K interactions in the environment, and is finally captured by the Rx. the frequency of the n-th subcarrier,and ∆f = W/N is the subcarrier spacing. Based on RT, we can identify the propagation paths between the transmitter and receiver. The path delay τi can be directly calculated from the total path propagation distance … view at source ↗
Figure 3
Figure 3. Figure 3: Uncertainty-aware EM property field calibration framework based on differentiable RT. The BCM provides Rx samples, inferred channel parameters, and uncertainty estimates for training. After calibration, the probabilistic supervision and gradient feedback represented by the dashed branch are removed, and the remaining forward pipeline is used for CSI computation in the WEDT. inferred values provided by the … view at source ↗
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visualization of channel gain across different datasets for WEDT and various baselines. Scene 1 and Scene 2 correspond to the indoor apartment and outdoor Munich scenes in the public dataset, respectively, while Scene 3 is a real-world corridor scene from Wireless Valley. The fixed TX location is marked with a red asterisk, and sparse training samples (M=30) are randomly distributed across each scene. The … view at source ↗
Figure 6
Figure 6. Figure 6: Visualization of CIR across different datasets for WEDT and various baselines. The three rows sequentially correspond to the three scenes. The subfigures from left to right in each row display the predicted CIRs at the locations marked by the numbered orange dots (positions 1, 2, and 3) shown in the top views of [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Impact of the training sample size on the D2D generalization performance of WEDT in Scene 1. The number of B2D training samples is set to 10, 30, 100, and 300. The left y-axis corresponds to MALE, while the right y-axis corresponds to SSIM and MCS. Offline Online Syn Mode Types 1.4 1.6 1.8 2.0 2.2 2.4 2.6 MALE Value 1.48 1.77 2.34 MALE (Left Axis) SSIM (Right Axis) MCS (Right Axis) 0.70 0.75 0.80 0.85 0.90… view at source ↗
Figure 9
Figure 9. Figure 9: Visualization of material-related spatial variations captured by the calibrated EM property field in Scene 3. (a) Real view of the selected wall. The red box indicates the metal door handle, and the yellow box indicates the transition region between the wooden door and the glass wall. (b) Reflection coefficient of the EM property field with object category priors. (c) Reflection coefficient of the EM prope… view at source ↗
Figure 12
Figure 12. Figure 12: CDF of the ALE for channel gain prediction achieved by RadioFlow trained on synthetic data generated based on WEDT and ITU labeling model. The test set is the D2D measurement dataset in Scene 3 and the marked points indicate the median ALE [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗
Figure 11
Figure 11. Figure 11: Coverage enhancement gain after RIS deployment under measurements, WEDT simulation, and ITU labeling simulation. The solid lines are the average gain over the target area at each RIS position. The dashed lines are the average gain across all RIS positions for each method. absolute gain. In measurements, the average RIS deployment gain across eight locations was 2.15 dB, whereas the ITU￾assigned model yiel… view at source ↗
read the original abstract

Digital twins (DTs) are promising for wireless deployment, optimization, and data generation, but building a propagation-faithful twin from sparse real measurements remains difficult. This paper proposes a wireless environment digital twin (WEDT) construction paradigm that evolves a reconstructed geometric DT into a propagation-consistent wireless environment representation through calibration of a scene-level electromagnetic (EM) property field. Instead of directly fitting link-specific channel responses, the proposed paradigm first constructs a geometry-prior Bayesian channel map (BCM) to convert sparse position-labeled channel state information (CSI) into dense probabilistic supervision with uncertainty estimates. It then embeds the learnable EM property field into differentiable ray tracing (RT) based channel computation, thereby enabling calibration through an explicit propagation chain. Experiments in both public and real-world scenes show that WEDT achieves accurate channel prediction, generalizes to unseen transceiver topologies, and remains effective across different sampling conditions. WEDT also offers utility for material-related environment sensing, more reliable physical-layer planning, and higher-quality synthetic data generation for wireless AI. These results demonstrate the value of the proposed paradigm for propagation-consistent WEDT construction and related wireless applications.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a wireless environment digital twin (WEDT) construction paradigm that starts from a reconstructed geometric DT and evolves it into a propagation-consistent representation by calibrating a scene-level electromagnetic (EM) property field. Sparse position-labeled CSI is first converted into dense probabilistic supervision via a geometry-prior Bayesian channel map (BCM) with uncertainty estimates; this supervision then drives calibration inside a differentiable ray-tracing (RT) pipeline. Experiments on public and real-world scenes report accurate channel prediction, generalization to unseen transceiver topologies, and utility for material sensing and synthetic data generation.

Significance. If the central claim holds, the work would provide a practical route to propagation-faithful DTs from sparse measurements, directly supporting physical-layer planning and higher-quality synthetic data for wireless AI. The explicit embedding of a learnable EM field into differentiable RT and the use of BCM for uncertainty-aware supervision are genuine strengths that distinguish the approach from direct channel fitting.

major comments (2)
  1. The central claim rests on the premise that a pre-reconstructed geometric DT is accurate enough that calibrating only the scene-level EM property field suffices. No sensitivity analysis or ablation is shown for the effect of geometry errors (object positions, surface normals, or ray-path inaccuracies) on the resulting multipath geometry and delay spread; if this premise fails, permittivity/conductivity adjustments alone cannot recover correct propagation.
  2. Experiments section: while positive results are claimed for public and real-world scenes, the manuscript does not report error bars, explicit baseline comparisons, or data-exclusion rules. This makes it difficult to evaluate the reported generalization to unseen transceiver topologies and effectiveness across sampling conditions.
minor comments (2)
  1. Notation for the BCM and the EM property field should be introduced with explicit definitions and dimensions in the method section to improve readability.
  2. Figure captions for the ray-tracing pipeline and BCM visualization would benefit from clearer labeling of input CSI locations versus predicted paths.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and positive assessment of the significance of our work. We address each major comment below and commit to revisions that directly respond to the concerns raised.

read point-by-point responses

  1. Referee: The central claim rests on the premise that a pre-reconstructed geometric DT is accurate enough that calibrating only the scene-level EM property field suffices. No sensitivity analysis or ablation is shown for the effect of geometry errors (object positions, surface normals, or ray-path inaccuracies) on the resulting multipath geometry and delay spread; if this premise fails, permittivity/conductivity adjustments alone cannot recover correct propagation.

    Authors: We agree that this is a key assumption and that the manuscript would be strengthened by explicit analysis. The proposed method takes a reconstructed geometric DT as given input (as is standard when geometry is obtained via separate sensing modalities such as LiDAR or photogrammetry). In the revision we will add a sensitivity study that introduces controlled perturbations to object positions, surface normals, and ray-path selection, then reports the resulting degradation in channel prediction accuracy and delay spread. This will clarify the operating regime in which EM-field calibration alone is sufficient. revision: yes

  2. Referee: Experiments section: while positive results are claimed for public and real-world scenes, the manuscript does not report error bars, explicit baseline comparisons, or data-exclusion rules. This makes it difficult to evaluate the reported generalization to unseen transceiver topologies and effectiveness across sampling conditions.

    Authors: We acknowledge the need for greater statistical transparency. In the revised manuscript we will (i) report mean and standard-deviation error bars for all quantitative metrics across multiple random seeds or cross-validation folds, (ii) add explicit comparisons against relevant baselines including direct neural-network channel fitting and non-calibrated ray-tracing, and (iii) provide a clear description of the train/test splits, sampling-density conditions, and any data-exclusion criteria used for the generalization experiments. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation relies on physics-based calibration and external validation

full rationale

The described chain first converts sparse CSI into dense probabilistic supervision via a geometry-prior BCM, then calibrates a scene-level EM property field inside differentiable ray tracing to produce propagation-consistent predictions. This is a standard data-driven calibration procedure whose outputs for unseen transceiver topologies are not equivalent to the inputs by construction; the RT propagation model supplies the extrapolation mechanism. Experiments on public and real-world scenes test generalization, confirming the method is externally falsifiable rather than self-referential. No equations or steps reduce a claimed prediction to a fitted input or self-citation by definition. The initial geometric DT accuracy is an explicit modeling assumption, not a circular loop.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 1 invented entities

The central claim depends on the geometric DT being a usable prior, the BCM producing reliable dense supervision from sparse CSI, and the differentiable RT engine being able to propagate gradients through the EM property field; these are not supplied by upstream literature but introduced or assumed for this work.

free parameters (1)
  • scene-level EM property field
    Learnable parameters of the electromagnetic property field that are calibrated to match the BCM supervision.
axioms (1)
  • domain assumption A reconstructed geometric DT provides a sufficient structural prior for propagation modeling.
    Invoked when the paper states the geometric DT is evolved into a propagation-consistent representation solely by EM-field calibration.
invented entities (1)
  • scene-level electromagnetic (EM) property field no independent evidence
    purpose: To serve as the adjustable parameter that makes the ray-traced channels match the Bayesian channel map.
    New field introduced in the paper; no independent evidence outside the calibration objective is provided.

pith-pipeline@v0.9.0 · 5757 in / 1392 out tokens · 38957 ms · 2026-05-22T03:55:07.805276+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

48 extracted references · 48 canonical work pages · 1 internal anchor

  1. [1]

    What should 6g be?

    S. Dang, O. Amin, B. Shihada, and M.-S. Alouini, “What should 6g be?”Nature Electronics, vol. 3, no. 1, pp. 20–29, 2020

    work page 2020

  2. [2]

    Wireless communications with programmable metasurface: New paradigms, opportunities, and challenges on transceiver design,

    W. Tang, M. Z. Chen, J. Y . Dai, Y . Zeng, X. Zhao, S. Jin, Q. Cheng, and T. J. Cui, “Wireless communications with programmable metasurface: New paradigms, opportunities, and challenges on transceiver design,” IEEE Wireless Communications, vol. 27, no. 2, pp. 180–187, 2020

    work page 2020

  3. [3]

    Integrated sensing and communications: Toward dual-functional wire- less networks for 6g and beyond,

    F. Liu, Y . Cui, C. Masouros, J. Xu, T. X. Han, Y . C. Eldar, and S. Buzzi, “Integrated sensing and communications: Toward dual-functional wire- less networks for 6g and beyond,”IEEE journal on selected areas in communications, vol. 40, no. 6, pp. 1728–1767, 2022

    work page 2022

  4. [4]

    Large ai models for wireless physical layer,

    J. Guo, Y . Cui, S. Jin, and J. Zhang, “Large ai models for wireless physical layer,”IEEE Communications Magazine, 2026

    work page 2026

  5. [5]

    Twelve scientific challenges for 6g: Rethinking the foundations of communications the- ory,

    M. Chafii, L. Bariah, S. Muhaidat, and M. Debbah, “Twelve scientific challenges for 6g: Rethinking the foundations of communications the- ory,”IEEE Communications Surveys & Tutorials, vol. 25, no. 2, pp. 868–904, 2023

    work page 2023

  6. [6]

    Nine challenges in artificial intelligence and wireless communications for 6g,

    W. Tong and G. Y . Li, “Nine challenges in artificial intelligence and wireless communications for 6g,”IEEE Wireless Communications, vol. 29, no. 4, pp. 140–145, 2022

    work page 2022

  7. [7]

    Overview of ai and communication for 6g network: Fundamentals, challenges, and future research opportunities,

    Q. Cuiet al., “Overview of ai and communication for 6g network: Fundamentals, challenges, and future research opportunities,”Science China Information Sciences, vol. 68, no. 7, p. 171301, 2025

    work page 2025

  8. [8]

    Deploy- ment design for multi-uav-assisted iot networks: A digital twin-driven deep reinforcement learning approach,

    L. Zhao, Z. Fei, J. Huang, X. Wang, B. Li, and W. Yuan, “Deploy- ment design for multi-uav-assisted iot networks: A digital twin-driven deep reinforcement learning approach,”IEEE Transactions on Wireless Communications, 2025

    work page 2025

  9. [9]

    Digital twin aided massive mimo csi feedback: Exploring the impact of twinning fidelity,

    H. Luo, S. Jiang, S. R. Khosravirad, and A. Alkhateeb, “Digital twin aided massive mimo csi feedback: Exploring the impact of twinning fidelity,”IEEE Transactions on Communications, 2025

    work page 2025

  10. [10]

    Radarsplat: Radar gaussian splatting for high-fidelity data synthesis and 3d reconstruction of autonomous driving scenes,

    P.-C. Kunget al., “Radarsplat: Radar gaussian splatting for high-fidelity data synthesis and 3d reconstruction of autonomous driving scenes,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2025, pp. 27 596–27 606

    work page 2025

  11. [11]

    Vggt: Visual geometry grounded transformer,

    J. Wang, M. Chen, N. Karaev, A. Vedaldi, C. Rupprecht, and D. Novotny, “Vggt: Visual geometry grounded transformer,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 5294– 5306

    work page 2025

  12. [12]

    Wireless InSite

    Remcom. Wireless InSite. [Online]. Available: https://www.remcom. com/wireless-insite-em-propagation-software

    work page

  13. [13]

    Sionna rt: Differentiable ray tracing for radio propagation modeling,

    J. Hoydis, F. A ¨ıt Aoudia, S. Cammerer, M. Nimier-David, N. Binder, G. Marcus, and A. Keller, “Sionna rt: Differentiable ray tracing for radio propagation modeling,” in2023 IEEE Globecom Workshops (GC Wkshps). IEEE, 2023, pp. 317–321

    work page 2023

  14. [14]

    Machine learning-assisted calibration for ray-tracing channel simulation at centimeter-wave and millimeter-wave bands,

    C. Yi, W. Chen, Q. Wu, and H. Wang, “Machine learning-assisted calibration for ray-tracing channel simulation at centimeter-wave and millimeter-wave bands,”IEEE Antennas and Wireless Propagation Let- ters, vol. 23, no. 5, pp. 1623–1627, 2024

    work page 2024

  15. [15]

    Recommendation ITU-R P.2040-2: Effects of Building Materi- als and Structures on Radiowave Propagation Above About 100 MHz,

    ITU-R, “Recommendation ITU-R P.2040-2: Effects of Building Materi- als and Structures on Radiowave Propagation Above About 100 MHz,” Recommendation ITU-R P.2040-2, Sep. 2021, [Online]. Available: https: //www.itu.int/rec/R-REC-P.2040-2-202109-I

    work page 2040

  16. [16]

    Boston twin: the boston digital twin for ray-tracing in 6g networks,

    P. Testolina, M. Polese, P. Johari, and T. Melodia, “Boston twin: the boston digital twin for ray-tracing in 6g networks,” inProceedings of the 15th ACM Multimedia Systems Conference, 2024, pp. 441–447

    work page 2024

  17. [17]

    3d wi-fi signal measurement in realistic digital twin testbed environments using ray tracing,

    M. Wang, H. Wang, H. Dong, and A. El Saddik, “3d wi-fi signal measurement in realistic digital twin testbed environments using ray tracing,”IEEE Transactions on Instrumentation and Measurement, 2026

    work page 2026

  18. [18]

    How to Bridge the Sim-to-Real Gap in Digital Twin-Aided Telecommunication Networks

    C. Ruah, H. Sifaou, O. Simeone, and B. M. Al-Hashimi, “How to bridge the sim-to-real gap in digital twin-aided telecommunication networks,” arXiv preprint arXiv:2507.07067, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  19. [19]

    Calibration of nyuray, a 3d mmwave and sub-thz ray tracer using indoor, outdoor, and factory channel measurements,

    O. Kanhere and T. S. Rappaport, “Calibration of nyuray, a 3d mmwave and sub-thz ray tracer using indoor, outdoor, and factory channel measurements,” inICC 2023-IEEE International Conference on Com- munications. IEEE, 2023, pp. 5054–5059

    work page 2023

  20. [20]

    Learning radio environments by dif- ferentiable ray tracing,

    J. Hoydis, F. A. Aoudia, S. Cammerer, F. Euchner, M. Nimier-David, S. T. Brink, and A. Keller, “Learning radio environments by dif- ferentiable ray tracing,”IEEE Transactions on Machine Learning in Communications and Networking, vol. 2, pp. 1527–1539, 2024

    work page 2024

  21. [21]

    Learnable wireless digital twins: Reconstructing elec- tromagnetic field with neural representations,

    S. Jianget al., “Learnable wireless digital twins: Reconstructing elec- tromagnetic field with neural representations,”IEEE Open Journal of the Communications Society, vol. 6, pp. 1568–1590, 2025

    work page 2025

  22. [22]

    Digital twin enabled simultaneous learning and modeling for uav-assisted secure wireless sensing in unknown environment,

    J. Yuan, L. Li, S. Gong, B. Gu, and F. Li, “Digital twin enabled simultaneous learning and modeling for uav-assisted secure wireless sensing in unknown environment,” inICC 2025-IEEE International Conference on Communications. IEEE, 2025, pp. 1524–1529

    work page 2025

  23. [23]

    Channel charting for uav navigation in ris-assisted isac systems,

    K. Sun, J. Zhang, C.-K. Wen, and S. Jin, “Channel charting for uav navigation in ris-assisted isac systems,”IEEE Journal of Selected Topics in Electromagnetics, Antennas and Propagation, 2025

    work page 2025

  24. [24]

    Digital twin of channel: Diffusion model for sensing-assisted statistical channel state information generation,

    X. Gong, X. Liu, A.-A. Lu, X. Gao, X.-G. Xia, C.-X. Wang, and X. You, “Digital twin of channel: Diffusion model for sensing-assisted statistical channel state information generation,”IEEE Transactions on Wireless Communications, 2025

    work page 2025

  25. [25]

    Ai-based csi feedback with digital twins: Real-world validation and insights,

    T.-H. Huang, C.-K. Wen, S.-H. Tsai, and T. Q. Duong, “Ai-based csi feedback with digital twins: Real-world validation and insights,”IEEE Wireless Communications Letters, 2025

    work page 2025

  26. [26]

    Digital twin-assisted explainable ai for robust beam prediction in mmwave mimo systems,

    N. Khan, A. Abdallah, A. Celik, A. M. Eltawil, and S. Coleri, “Digital twin-assisted explainable ai for robust beam prediction in mmwave mimo systems,”IEEE Transactions on Wireless Communications, 2025

    work page 2025

  27. [27]

    Nerf: Representing scenes as neural radiance fields for view synthesis,

    B. Mildenhallet al., “Nerf: Representing scenes as neural radiance fields for view synthesis,”Communications of the ACM, vol. 65, no. 1, pp. 99– 106, 2021

    work page 2021

  28. [28]

    3d gaussian splatting for real-time radiance field rendering

    B. Kerbl, G. Kopanas, T. Leimk ¨uhler, G. Drettakiset al., “3d gaussian splatting for real-time radiance field rendering.”ACM Trans. Graph., vol. 42, no. 4, pp. 139–1, 2023

    work page 2023

  29. [29]

    Nerf2: Neural radio-frequency radiance fields,

    X. Zhao, Z. An, Q. Pan, and L. Yang, “Nerf2: Neural radio-frequency radiance fields,” inProceedings of the 29th Annual International Con- ference on Mobile Computing and Networking, 2023, pp. 1–15

    work page 2023

  30. [30]

    R-nerf: Neural radiance fields for modeling ris-enabled wireless environments,

    H. Yang, Z. Jin, C. Wu, R. Xiong, R. C. Qiu, and Z. Ling, “R-nerf: Neural radiance fields for modeling ris-enabled wireless environments,” inGLOBECOM 2024-2024 IEEE Global Communications Conference. IEEE, 2024, pp. 3859–3864

    work page 2024

  31. [31]

    Wrf-gs: Wireless radiation field reconstruction with 3d gaussian splatting,

    C. Wen, J. Tong, Y . Hu, Z. Lin, and J. Zhang, “Wrf-gs: Wireless radiation field reconstruction with 3d gaussian splatting,” inIEEE INFOCOM 2025-IEEE Conference on Computer Communications. IEEE, 2025, pp. 1–10

    work page 2025

  32. [32]

    Rf-3dgs: Wireless channel modeling with radio radiance field and 3d gaussian splatting,

    L. Zhang, H. Sun, S. Berweger, C. Gentile, and R. Q. Hu, “Rf-3dgs: Wireless channel modeling with radio radiance field and 3d gaussian splatting,”IEEE Transactions on Wireless Communications, vol. 25, pp. 10 419–10 433, 2026

    work page 2026

  33. [33]

    Capability of 3-d ray tracing for defining parameter sets for the specification of future mobile communications systems,

    T. Fugen, J. Maurer, T. Kayser, and W. Wiesbeck, “Capability of 3-d ray tracing for defining parameter sets for the specification of future mobile communications systems,”IEEE Transactions on Antennas and Propagation, vol. 54, no. 11, pp. 3125–3137, 2006

    work page 2006

  34. [34]

    Study on channel model for frequencies from 0.5 to 100 ghz,

    3GPP, “Study on channel model for frequencies from 0.5 to 100 ghz,” 3rd Generation Partnership Project (3GPP), Tech. Rep. TR 38.901 V17.0.0, Mar. 2022

    work page 2022

  35. [35]

    Ray techniques in electromagnetics,

    G. A. Deschamps, “Ray techniques in electromagnetics,”Proceedings of the IEEE, vol. 60, no. 9, pp. 1022–1035, 2005

    work page 2005

  36. [36]

    Propagation by diffraction,

    ITU-R, “Propagation by diffraction,” International Telecommunication Union, Tech. Rep. ITU-R P.526-15, Oct. 2019

    work page 2019

  37. [37]

    Analysis and modeling on co-and cross-polarized urban radio propaga- tion for dual-polarized mimo wireless systems,

    V . Degli-Esposti, V .-M. Kolmonen, E. M. Vitucci, and P. Vainikainen, “Analysis and modeling on co-and cross-polarized urban radio propaga- tion for dual-polarized mimo wireless systems,”IEEE transactions on antennas and propagation, vol. 59, no. 11, pp. 4247–4256, 2011

    work page 2011

  38. [38]

    Sionna rt: Technical report,

    F. A. Aoudiaet al., “Sionna rt: Technical report,”arXiv preprint arXiv:2504.21719, 2025

    work page arXiv 2025

  39. [39]

    Prediction of wireless channel statistics with ray tracing and uncalibrated digital twin,

    M. S. Abouamer, R. J. Williams, and P. Popovski, “Prediction of wireless channel statistics with ray tracing and uncalibrated digital twin,”IEEE Wireless Communications Letters, 2025

    work page 2025

  40. [40]

    Goldsmith,Wireless communications

    A. Goldsmith,Wireless communications. Cambridge university press, 2005

    work page 2005

  41. [41]

    A new path- gain/delay-spread propagation model for digital cellular channels,

    L. J. Greenstein, V . Erceg, Y . S. Yeh, and M. V . Clark, “A new path- gain/delay-spread propagation model for digital cellular channels,”IEEE Transactions on Vehicular Technology, vol. 46, no. 2, pp. 477–485, 2002

    work page 2002

  42. [42]

    Bayesian active learning for sample efficient 5g radio map reconstruction,

    K. D. Polyzoset al., “Bayesian active learning for sample efficient 5g radio map reconstruction,”IEEE Transactions on Wireless Communica- tions, vol. 23, no. 12, pp. 19 382–19 396, 2024

    work page 2024

  43. [43]

    Radio environment map construction based on gaussian process with positional uncertainty,

    P. Zhenet al., “Radio environment map construction based on gaussian process with positional uncertainty,”IEEE Wireless Communications Letters, vol. 11, no. 8, pp. 1639–1643, 2022

    work page 2022

  44. [44]

    Colored point cloud registration revisited,

    J. Park, Q.-Y . Zhou, and V . Koltun, “Colored point cloud registration revisited,” inICCV, 2017

    work page 2017

  45. [45]

    Openstreetmap,

    OpenStreetMap Foundation, “Openstreetmap,” https://www. openstreetmap.org, 2023

    work page 2023

  46. [46]

    3D Scanner App,

    Laan Labs, “3D Scanner App,” https://3dscannerapp.com/, accessed: 2026-04-12

    work page 2026

  47. [47]

    Multi-state ris-enabled near-field separate channel es- timation through multi-carrier nomp,

    X. Yeet al., “Multi-state ris-enabled near-field separate channel es- timation through multi-carrier nomp,” in2025 Seventeenth Interna- tional Conference on Wireless Communications and Signal Processing (WCSP). IEEE, 2025, pp. 1–6

    work page 2025

  48. [48]

    Radioflow: Efficient radio map construction framework with flow matching,

    H. Jiaet al., “Radioflow: Efficient radio map construction framework with flow matching,”arXiv preprint arXiv:2510.09314, 2025

    work page arXiv 2025