Multi-Modal Neural Radio Radiance Field for Localized Statistical Channel Modelling

Shutao Zhang; Tsung-Hui Chang; Ye Xue; Yiheng Wang

REVIEW 5 major objections 6 minor 17 references

Multi-Modal Neural Radio Radiance Field for Localized Statistical Channel Modelling

T0 review · 5 major / 6 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A self-supervised volume-rendering network fuses LiDAR and RSRP to reconstruct channel angular power spectra across a coverage area, including unmeasured locations, and outperforms RSRP-only compressed sensing.

desk verdict A promising NeRF-based channel modeling idea whose forward model is not derived and whose experiments are synthetic; deserves a serious referee but needs major revision. read the letter →

arxiv 2508.06054 v1 pith:QJKH4ZSR submitted 2025-08-08 eess.SP

Yiheng Wang , Shutao Zhang , Ye Xue , Tsung-Hui Chang This is my paper · ORCID

classification eess.SP

keywords localizedstatisticalchannelmodelingangularpowerspectrumneuralradiancefieldLiDARpointcloudRSRPself-supervisedlearningvolumerenderingsphericalharmonics

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

MM-LSCM claims that fusing LiDAR point clouds with beam-wise RSRP measurements in a self-supervised neural radiance field reconstructs the channel angular power spectrum (APS) across an entire coverage area, including locations never measured. Existing LSCM treats each grid as a standalone compressed-sensing problem, so unvisited grids cannot be predicted and environmental structure is ignored. The paper's network renders the APS by volume integration: each voxel emits a learned direction-dependent complex signal, and an environment-informed stopping probability weights which voxels actually contribute. Compared with the RSRP-only baseline, the method cuts MAE from 1.85 dB to 0.87 dB on measured grids and reaches 0.56 to 0.90 dB on unmeasured grids, and it tolerates 3 dB noise on both measurement matrix and RSRP. If this holds, operators could replace expensive drive tests with sparse RSRP plus LiDAR sweeps.

What carries the argument

The load-bearing object is the stopping probability $\nu_n(t) = T_n(0,t)\,\sigma_n(t)$, where $\sigma_n(t)$ is the network-predicted hit probability at a voxel and $T_n(0,t)=\exp(-\int_0^t \sigma_n(t')\,dt')$ is the pass probability. This one weight aligns the radio and environment modalities: it controls how much each voxel's radiated signal contributes to the synthesized APS, and it also renders the obstacle depth used as a geometric self-supervision signal. The APS element for direction $n$ is formed as $|\int_0^{t_{\max}}\nu_n(t)\,S_{l,n}(t)\,dt|^2$, with $S_{l,n}$ the learned spherical-harmonics radiance from the voxel toward grid $l$; the whole pipeline is a ray-marching approximation

What would settle it

Take a synthetic scene whose true APS is a few known point sources generated from Eq. (1), train MM-LSCM on RSRP produced by Eq. (21), and test whether the recovered APS localizes those source angles and gains with small error. If the model matches the RSRP but assigns energy to wrong angles, or if its APS estimate shifts when voxel resolution or ray discretization changes, the volume-rendering representation is not faithful to the channel model.

Watch

Extended reading notes

Core claim

The central claim is that environmental geometry, injected as per-voxel density from a LiDAR point cloud, supplies the missing spatial prior for localized statistical channel modeling. The proposed network learns two coupled fields along rays from the base station: a hitting probability (volume density) that encodes where the signal is blocked or scattered, and a direction-dependent complex radiance represented by spherical harmonics. Volume rendering combines these into a complex gain for each discrete angle, whose squared magnitude is the APS; the same rendering, weighted by a stopping probability, also produces an expected obstacle depth. Training is self-supervised: on measured grids the

Load-bearing premise

The method assumes the volume-rendered integral of per-voxel radiance times stopping probability faithfully represents the discrete, sparse multipath angular power spectrum of Eq. (1); no derivation links the two, so if the mapping is unfaithful even unlimited training data cannot recover the true APS.

Editorial extensions

If this is right

Operators can estimate full angular power spectra across a coverage area from sparse drive-test RSRP plus LiDAR, without labeled channel data.
Unvisited grids become predictable, so network optimization can simulate changes in antenna tilt and azimuth in areas never measured.
The method keeps its accuracy when both the codebook response and the RSRP measurements are noisy, which matches real-world conditions.
Removing the environment branch, as in the single-modality ablation, raises MAE and increases MSE by orders of magnitude on the measured-grid task, so the LiDAR prior is doing real work.
The depth-rendering side output means the model produces a coarse environmental map as a byproduct, potentially reusable for site planning.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Since the performance gain comes from injecting geometry as a prior, the MAE gap over RSRP-only solvers should widen in cluttered environments and nearly vanish in open areas; this is testable by re-running the same pipeline on synthetic scenes with controlled scatterer density.
The stopping probability is a single shared weight for radio contribution and physical depth; this suggests using predicted depth in unexplored regions as an audit signal to check whether the model's radio predictions are grounded in real geometry rather than fitted to the training RSRP.
The architecture only needs a per-voxel density prior, so the same self-supervised scheme could ingest other sensors, such as cameras, coarse ray-traced maps, or building footprints, without changing the rendering equations.
Because the learned field is continuous, one could simulate antenna rotations or relocations by re-evaluating the measurement matrix against the same radiance field, enabling rapid what-if network planning without new measurements.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Haibo, this one is worth a look, with caveats. The core idea — using NeRF-style volume rendering to turn sparse RSRP plus LiDAR into a full angular power spectrum, and to extrapolate to unexplored regions — is a real and potentially useful step for radio map estimation. The architecture is a sensible adaptation of NeRF2: a dual-branch network where LiDAR density conditions the attenuation branch, and a spherical-harmonics radiance branch produces complex signals. The stopping-probability coherence weight that links radio rendering and depth prediction is a nice touch, and the self-supervised loss over RSRP is a natural way to avoid labeled APS.

What the paper actually does well is give a concrete trainable model and show, on the NeRF2 dataset, that it beats the two implemented comparisons (WNOMP and a LiDAR-free ablation) on both explored and unexplored grids. The gap on unexplored regions is the most interesting result.

Now the soft spots. They are not cosmetic. Eq. (5) defines the APS as a complex expectation of the channel gains, which is inconsistent with Eq. (4) and (6) where it must be the power spectrum E|α|^2. That looks like a dropped absolute square, but it is in the formal model and needs fixing. More importantly, the volume-rendered integral in Eqs. (12)–(13) is asserted as the APS without any derivation connecting it to the discrete path model in Eq. (1). The claim that it follows from Huygens-Fresnel is not a derivation, and it matters because the entire predictive power of the model rests on this representation being faithful. The experiments are also synthetic: RSRP is generated as y=Φx from the ground-truth APS, not measured. That is acceptable for a proof of concept, but the robustness test then adds noise to this synthetic data and the already-known matrix, which is weaker evidence than real-world noisy RSRP. Finally, the paper gives no code, no hyperparameters, no error bars, and no comparison to simpler interpolation or Gaussian-process baselines. The WNOMP comparison is only fair on Sub-Task 1 because WNOMP is per-grid, but a stronger extrapolation baseline would make the claim more convincing.

All that said, the central hypothesis — that geometry from LiDAR improves channel extrapolation — is plausible, and the paper is honest about its limitations. It is not a finished result, but it deserves a serious referee. An editor should send it out with the expectation of major revision: clean up Eq. (5), add a real derivation or clear argument for the volume-to-APS mapping, release code and settings, and add error bars. If the forward model can be justified, this would be a solid journal paper. For now, read it to track the ideas, but don't cite it yet; wait for the revision.

Referee Report

5 major / 6 minor

Summary. The paper proposes MM-LSCM, a dual-branch neural network that combines beam-wise RSRP and LiDAR point clouds to predict angular power spectra (APS) over a coverage area. A NeRF-style volume rendering representation is used: each voxel is assigned a hitting probability and a direction-dependent complex radiance, and integration along AoD rays yields an APS estimate. Training is self-supervised through an RSRP reconstruction loss, with an auxiliary depth loss supervised by LiDAR occupancy. Experiments on the NeRF2 indoor dataset, using synthetic RSRP generated from measured spatial spectra, report lower MAE/MSE than WNOMP and a single-modality ablation.

Significance. The self-supervised formulation and the use of LiDAR as a geometric prior are genuinely attractive: they avoid labeled APS data and are aimed at predicting unexplored regions, which per-grid compressed-sensing baselines cannot do. The depth-supervision idea, tying the radio stopping probability to LiDAR occupancy, is a novel and reasonable soft-alignment mechanism. If the formal model and evaluation are made sound, the framework would be a useful contribution to data-driven localized channel modeling. However, the current manuscript does not yet establish the claimed reconstruction accuracy because several load-bearing definitions and mappings are unproven or inconsistent.

major comments (5)

[Sec. II, Eq. (5)] Equation (5) defines the APS as x_l = E_T[alpha_{l,t}] with alpha_{l,t} the complex path-gain vector. Under the paper's own uniform-phase assumption, E_T[alpha_{l,t}] = 0, and the quantity is complex, not a real nonnegative APS. Equations (4) and (6), and the loss (17), require x_l = E_T[|alpha_{n,l,t}|^2]. This is not a notational slip: as written, the formal model predicts zero RSRP. Please correct the definition and give the short derivation from Eq. (3) rather than citing [6] alone.
[Sec. III-D, Eqs. (12)-(13)] The mapping from the volume-rendered integral hat_r_{l,n} = integral nu_n(t) S_{l,n}(t) dt to the discrete sparse APS in Eq. (1) is asserted, not derived. Since sigma_n(t) and S_{l,n}(t) are unconstrained network outputs, minimizing (17) can fit Phi hat_x_l to y_l without recovering the true APS. Please provide a representational argument, or clearly state a modeling assumption under which the integral equals the desired path sum, or add a synthetic experiment with known sparse APS to demonstrate that hat_x_l is the underlying APS rather than a curve fit.
[Sec. IV-A, Eqs. (21)-(22)] The RSRP inputs are not measured. They are generated as y_l = Phi x_l and y_l^r = Phi_r x_l from the downsampled ground-truth spatial spectrum. Thus the abstract's 'real-world dataset' claim applies only to LiDAR and spatial spectrum; all radio training targets are synthetic and noiseless. Moreover, MSE in Eq. (23) is evaluated against exactly these x_l, so the comparison rewards models that invert the same generative model used to create the data. Please present real RSRP measurements, or explicitly reframe the study as a controlled synthetic benchmark and discuss the conditions under which the conclusions transfer.
[Sec. III-E, Eqs. (18)-(19)] Using the first nonzero-density voxel along each ray as the depth supervision target for the radio stopping probability is a strong physical assumption. In a multipath environment, the dominant contribution to a given AoD can arrive from a reflection beyond the first occupied voxel; forcing nu_n(t) to concentrate at t_{d*} will suppress legitimate later paths and bias hat_x_l. The paper provides no evidence that first-hit depth matches the radio-path depth. Please justify this assumption or relax it, for example by supervising only LoS rays.
[Sec. IV-C] The robustness experiment is not fully specified. 'Introduce 3 dB noise to both the measurement matrix and the RSRP data' could mean noise is added to the Phi used in training, to the Phi used at test time, or to the RSRP only. If the model is trained with clean Phi but evaluated with noisy Phi, the degradation may be an artifact of train/test mismatch; if noisy Phi is used in the loss (17), the forward model itself changes. State precisely how noise is added, at which stage, and how the WNOMP baseline is treated under the same perturbation.

minor comments (6)

[Sec. IV-A] The APS dimension is inconsistent: 'N=16201' and 'N=1620' appear in the same paragraph, while the footnote states 90 x 18 = 1620.
[Sec. II, Eq. (4)] There is an unmatched parenthesis: '|W^H S|^2)' should presumably be '(|W^H S|^2)'.
[Sec. III-C, Eq. (8)] The spherical-harmonics notation sh(tau, mu) is not defined. Please specify how tau encodes the SH coefficients and how mu is evaluated with the basis.
[Sec. IV-B, Eq. (23)] The MAE formula appears to be a sum over M beams, not a mean. Clarify whether the reported numbers are averaged over the M beams.
[Tables I and II] MSE values differ by orders of magnitude across methods (e.g., 0.0685 vs 153.93). State the normalization/scaling of x_l so the reader can interpret the comparison.
[Sec. IV-A] Clarify that the receiver in the NeRF2 dataset is treated as the BS while the original setup uses a fixed receiver and a moving tag; the coordinate transformations and codebook orientation used in the experiments should be specified.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the imported Eq. (4) model is a self-citation but independently derivable, and the training losses are self-supervised rather than fitted to test targets.

full rationale

The paper's central derivation is not circular. The only load-bearing imported relation is the RSRP-to-APS linear model in Eq. (4)/(6), cited as 'As derived in [6]' to the authors' own prior work. Although this is a self-citation with overlapping authors, it is a parameter-free statistical identity that follows directly from Eqs. (1)-(3) plus the stated independence and uniform-phase assumptions; it does not encode any output of the proposed MM-LSCM network and is therefore independent support for the observation model. The volume-rendering synthesis in Eqs. (12)-(13) is an architectural ansatz connecting voxel properties to APS, not a quantity fitted against the held-out APS or RSRP test targets. The training losses are self-supervised: Eq. (17) fits the observed RSRP through the forward model with an ℓ1 prior, and Eq. (19) is a self-consistency constraint between LiDAR-derived density and rendered depth. Neither uses ground-truth APS as supervision. The reported gains over WNOMP and SM-LSCM are therefore not forced by construction. The paper does contain a separate formal correctness risk: Eq. (5) writes x_l = E_T[α_{l,t}] (a complex expectation) where the surrounding model requires a nonnegative power vector E_T|α_{l,t}|^2, and Sections III-C/D do not rigorously connect the volume-rendered integral to the discrete path sum of Eq. (1). These are mathematical/representational concerns, not circularity, so they do not raise the circularity score.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim rests on the self-cited linear RSRP model, the Huygens-Fresnel volume-rendering analogy, the uniform-phase path-gain assumption, a sparsity prior, and an ad hoc LiDAR-depth coupling. Several hyperparameters that would be needed to reproduce the method (λ1, λ2, V, voxel sizes, D) are unreported. No genuinely new physical entities are introduced.

free parameters (6)

λ1
Sparsity regularization weight in the radio loss (Eq. 17); not reported, controls the sparsity of the recovered APS.
λ2
Environment loss weight in the total objective (Eq. 20); not reported, balances radio and depth supervision.
V
Position encoding order in γ(·) (Eq. 7, ref [15]); affects spatial resolution, not specified.
voxel resolutions Δx, Δy, Δz
Define the density tensor ρ (Sec. III-B) and therefore the depth supervision targets; not specified.
D
Number of ray segments for discrete volume rendering (Eq. 15); controls integration accuracy, not specified.
codebook shift (Δθ, Δφ) = (5°, 4°)
Chosen to simulate antenna rotation in the multi-round scenario (Sec. IV-A); experimental parameter.

assumptions (6)

domain assumption The linear RSRP model E_T[rsrp] = |W^H S|^2 x_l (Eq. 4) with x_l the channel APS, as derived in [6].
The entire measurement model and loss depend on this linear relation; Eq. 4 is cited to a prior paper by the same group, and the APS definition in Eq. 5 is inconsistent with it.
domain assumption Each voxel re-radiates the incident RF signal according to the Huygens-Fresnel principle (Sec. III-C).
This justifies the volume-rendering synthesis but no quantitative link to the discrete path model of Eq. (1) is provided.
domain assumption The complex path gains α have phases uniform on [−π,π] and are independent across paths (Sec. II).
The linear expectation in Eq. 4 relies on this; it also implies E[α]=0, contradicting Eq. (5)'s definition of x_l.
ad hoc to paper The first nonzero-density voxel along a ray is the correct depth target for the radio stopping probability (Eqs. 18-19).
This couples LiDAR geometry to radio attenuation without empirical or physical justification.
domain assumption The APS x_l is sparse (Sec. II, [6]).
Sparsity justifies the ℓ1 penalty and the compressed-sensing-style recovery.
domain assumption In the main evaluation, the measurement matrix Φ is exactly known and RSRP is generated noiselessly as y_l = Φx_l (Eq. 21).
The reported real-world RSRP is actually synthesized; the robustness test adds noise, but the main comparison uses clean synthetic measurements.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Multi-Modal Neural Radio Radiance Field for Localized Statistical Channel Modelling." pith.science (2026). https://pith.science/paper/QJKH4ZSR

@misc{pith2026250806054,
  author       = {Pith},
  title        = {Pith review of: Multi-Modal Neural Radio Radiance Field for Localized Statistical Channel Modelling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QJKH4ZSR}},
  note         = {Machine review of arXiv:2508.06054}
}

read the original abstract

This paper presents MM-LSCM, a self-supervised multi-modal neural radio radiance field framework for localized statistical channel modeling (LSCM) for next-generation network optimization. Traditional LSCM methods rely solely on RSRP data, limiting their ability to model environmental structures that affect signal propagation. To address this, we propose a dual-branch neural architecture that integrates RSRP data and LiDAR point cloud information, enhancing spatial awareness and predictive accuracy. MM-LSCM leverages volume-rendering-based multi-modal synthesis to align radio propagation with environmental obstacles and employs a self-supervised training approach, eliminating the need for costly labeled data. Experimental results demonstrate that MM-LSCM significantly outperforms conventional methods in channel reconstruction accuracy and robustness to noise, making it a promising solution for real-world wireless network optimization.

Figures

Figures reproduced from arXiv: 2508.06054 by the authors.

**Figure 2.** Illustration of Architecture of the Multi-Modal Neural Radio Radiance Network [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Illustration of Volume-Rendering-based Multi-Modal [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Illustration of Point Cloud of Laboratory Environment. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 14 canonical work pages

[6]

A physics-based and data-driven approach for localized statistical channel modeling,

S. Zhang, X. Ning, X. Zheng, Q. Shi, T.-H. Chang, and Z.-Q. Luo, “A physics-based and data-driven approach for localized statistical channel modeling,” IEEE Trans. Wireless Commun. , vol. 23, no. 6., June. 2024

work page 2024
[1]

SRCON: A data-driven network performance simulator for real-world wireless networks,

Z.-Q. Luo, X. Zheng, D. L ´opez-P´erez, Q. Yan, X. Chen, N. Wang, Q. Shi, T.-H. Chang, and A. Garcia-Rodriguez, “SRCON: A data-driven network performance simulator for real-world wireless networks,” IEEE Commun. Mag. , vol. 61, no. 6, pp. 96–102, Jun. 2023

work page 2023
[2]

Minimization of drive tests solution in 3GPP,

W. A. Hapsari, A. Umesh, M. Iwamura, M. Tomala, B. Gyula, and B. Sebire, “Minimization of drive tests solution in 3GPP,” IEEE Com- mun. Mag., vol. 50, no. 6, pp. 28–36, 2012

work page 2012
[3]

Statistical channel modeling methods in wireless network twinning: current status and frontiers,

S. Zhang, Y . Xue, Q. Shi, and T. Chang, “Statistical channel modeling methods in wireless network twinning: current status and frontiers,” ZTE Technology Journal, vol. 29, no. 3, pp. 26–31, 2023

work page 2023
[4]

A survey of 5G channel measurements and models,

C.-X. Wang, J. Bian, J. Sun, W. Zhang, and M. Zhang, “A survey of 5G channel measurements and models,” IEEE Commun. Surveys Tuts. , vol. 20, no. 4, pp. 3142–3168, 2018

work page 2018
[5]

The design and applications of high-performance ray-tracing simulation platform for 5G and beyond wireless communications: A tutorial,

D. He, B. Ai, K. Guan, L. Wang, Z. Zhong, and T. K ¨urner, “The design and applications of high-performance ray-tracing simulation platform for 5G and beyond wireless communications: A tutorial,” IEEE Commun. Surveys Tuts., vol. 21, no. 1, pp. 10–27, 2018

work page 2018
[7]

Neural enhanced variational Bayesian inference on graphs for localized statis- tical channel modeling,

Y . Wang, S. Zhang, Y . Xue, T. Yu, Q. Shi, and T.-H. Chang, “Neural enhanced variational Bayesian inference on graphs for localized statis- tical channel modeling,” in Proc. IEEE Int. Conf. on Commun. (ICC) , 2024, pp. 342–347

work page 2024
[8]

Winert: Towards neural ray tracing for wireless channel modelling and differentiable simulations,

T. Orekondy, P. Kumar, S. Kadambi, H. Ye, J. Soriaga, and A. Behboodi, “Winert: Towards neural ray tracing for wireless channel modelling and differentiable simulations,” in The Eleventh Proc. Int. Conf. Learn. Represent., 2023

work page 2023

Show all 17 references

[9]

Fast and efficient cross band channel prediction using machine learning,

A. Bakshi, Y . Mao, K. Srinivasan, and S. Parthasarathy, “Fast and efficient cross band channel prediction using machine learning,” in The 25th Annual International Conference on Mobile Computing and Networking, 2019, pp. 1–16

2019
[10]

Nerf2: Neural radio-frequency radiance fields,

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

2023
[11]

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

B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, “Nerf: Representing scenes as neural radiance fields for view synthesis,” Commun. of the ACM , vol. 65, no. 1, pp. 99–106, 2021

2021
[12]

How to define the propagation environment semantics and its application in scatterer-based beam prediction,

Y . Sun, J. Zhang, L. Yu, Z. Zhang, and P. Zhang, “How to define the propagation environment semantics and its application in scatterer-based beam prediction,” IEEE Wireless Commun.Lett., vol. 12, no. 4, pp. 649– 653, 2023

2023
[13]

Multi-grid-based localized statistical channel modeling: A radio map approach,

X. Ning, S. Zhang, X. Zheng, and T.-H. Chang, “Multi-grid-based localized statistical channel modeling: A radio map approach,” in 2022 Int. Symp. on Wireless Commun. Syst. (ISWCS) . IEEE, 2022, pp. 1–6

2022
[14]

Tse and P

D. Tse and P. Viswanath, Fundamentals of wireless communication . Cambridge university press, 2005

2005
[15]

Fourier features let networks learn high frequency functions in low dimensional domains,

M. Tancik, P. Srinivasan, B. Mildenhall, S. Fridovich-Keil, N. Raghavan, U. Singhal, R. Ramamoorthi, J. Barron, and R. Ng, “Fourier features let networks learn high frequency functions in low dimensional domains,” Proc. Adv. Neural Inf. Process. Syst. , vol. 33, pp. 7537–7547, 2020

2020
[16]

Plenoxels: Radiance fields without neural networks,

S. Fridovich-Keil, A. Yu, M. Tancik, Q. Chen, B. Recht, and A. Kanazawa, “Plenoxels: Radiance fields without neural networks,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2022, pp. 5501–5510

2022
[17]

V olume rendering digest (for nerf),

A. Tagliasacchi and B. Mildenhall, “V olume rendering digest (for nerf),” arXiv preprint arXiv:2209.02417 , 2022

2022 arXiv

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [6]

A physics-based and data-driven approach for localized statistical channel modeling,

S. Zhang, X. Ning, X. Zheng, Q. Shi, T.-H. Chang, and Z.-Q. Luo, “A physics-based and data-driven approach for localized statistical channel modeling,” IEEE Trans. Wireless Commun. , vol. 23, no. 6., June. 2024

work page 2024

[2] [1]

SRCON: A data-driven network performance simulator for real-world wireless networks,

Z.-Q. Luo, X. Zheng, D. L ´opez-P´erez, Q. Yan, X. Chen, N. Wang, Q. Shi, T.-H. Chang, and A. Garcia-Rodriguez, “SRCON: A data-driven network performance simulator for real-world wireless networks,” IEEE Commun. Mag. , vol. 61, no. 6, pp. 96–102, Jun. 2023

work page 2023

[3] [2]

Minimization of drive tests solution in 3GPP,

W. A. Hapsari, A. Umesh, M. Iwamura, M. Tomala, B. Gyula, and B. Sebire, “Minimization of drive tests solution in 3GPP,” IEEE Com- mun. Mag., vol. 50, no. 6, pp. 28–36, 2012

work page 2012

[4] [3]

Statistical channel modeling methods in wireless network twinning: current status and frontiers,

S. Zhang, Y . Xue, Q. Shi, and T. Chang, “Statistical channel modeling methods in wireless network twinning: current status and frontiers,” ZTE Technology Journal, vol. 29, no. 3, pp. 26–31, 2023

work page 2023

[5] [4]

A survey of 5G channel measurements and models,

C.-X. Wang, J. Bian, J. Sun, W. Zhang, and M. Zhang, “A survey of 5G channel measurements and models,” IEEE Commun. Surveys Tuts. , vol. 20, no. 4, pp. 3142–3168, 2018

work page 2018

[6] [5]

The design and applications of high-performance ray-tracing simulation platform for 5G and beyond wireless communications: A tutorial,

D. He, B. Ai, K. Guan, L. Wang, Z. Zhong, and T. K ¨urner, “The design and applications of high-performance ray-tracing simulation platform for 5G and beyond wireless communications: A tutorial,” IEEE Commun. Surveys Tuts., vol. 21, no. 1, pp. 10–27, 2018

work page 2018

[7] [7]

Neural enhanced variational Bayesian inference on graphs for localized statis- tical channel modeling,

Y . Wang, S. Zhang, Y . Xue, T. Yu, Q. Shi, and T.-H. Chang, “Neural enhanced variational Bayesian inference on graphs for localized statis- tical channel modeling,” in Proc. IEEE Int. Conf. on Commun. (ICC) , 2024, pp. 342–347

work page 2024

[8] [8]

Winert: Towards neural ray tracing for wireless channel modelling and differentiable simulations,

T. Orekondy, P. Kumar, S. Kadambi, H. Ye, J. Soriaga, and A. Behboodi, “Winert: Towards neural ray tracing for wireless channel modelling and differentiable simulations,” in The Eleventh Proc. Int. Conf. Learn. Represent., 2023

work page 2023