pith. sign in

arxiv: 2604.20494 · v1 · submitted 2026-04-22 · 📡 eess.SP

Near-Field Wideband Channel Estimation for XL-MIMO Systems via Denoising Diffusion Model

Qingxia Feng , Yin Fang , Meng Hua , Cheng Zhang , Chunguo Li , Yongming Huang , Luxi Yang This is my paper

Pith reviewed 2026-05-10 00:02 UTC · model grok-4.3

classification 📡 eess.SP
keywords XL-MIMOwideband channel estimationdiffusion modelnear-fieldspatial non-stationaritybeam splitmulti-scale attentiondenoising network
0
0 comments X

The pith

A diffusion model learns the prior distribution of near-field wideband channels to solve XL-MIMO estimation as Bayesian inference.

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

The paper shows that near-field wideband channels in XL-MIMO systems exhibit spatial non-stationarity and beam split effects, which can be captured by training a diffusion model on the channel prior. Estimation is then performed by sampling from the learned posterior using a custom denoising network built with multi-scale attention. A reader would care because these effects grow severe with larger arrays and wider bandwidths, making traditional estimators unreliable for 6G links. If the claim holds, the approach supplies accurate channel knowledge where conventional methods degrade, supporting reliable beamforming and data transmission.

Core claim

The authors analyze statistical correlations to establish that near-field wideband channels display both spatial non-stationarity and beam split effects, then cast channel estimation as Bayesian posterior inference in which a diffusion model supplies the prior. They introduce a denoising network that extracts multi-scale spatial-frequency features through parallel convolutional branches of different receptive fields and fuses feature attention with spatial attention to emphasize critical structures, yielding more accurate modeling of the channel distribution.

What carries the argument

A denoising diffusion model whose prior is learned via a multi-scale attention network that runs parallel convolutional branches and combines feature and spatial attention modules.

Load-bearing premise

The diffusion model accurately captures the true prior distribution of real near-field wideband channels rather than merely fitting the simulated training data.

What would settle it

Measured channel data from a physical XL-MIMO testbed in which the proposed estimator produces equal or higher error than standard baselines such as compressed sensing or least-squares methods under matched conditions.

Figures

Figures reproduced from arXiv: 2604.20494 by Cheng Zhang, Chunguo Li, Luxi Yang, Meng Hua, Qingxia Feng, Yin Fang, Yongming Huang.

Figure 1
Figure 1. Figure 1: System model of the near-field wideband channel. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The spatial correlation value. The Pr(ψ) is modeled by the power delay profile [35], [36], given by Pr(ψ) = Kr exp  − ψ σψ  , ψ > 0, (20) where Kr = 1/σψ is a normalization factor and σψ denotes the standard deviation. Then, I a r = 1 1 + jσψ 2πfm(2n∆n+∆2 n)d2ξ¯0 cr¯0 , (21) and |Ia r | = 1 r 1 + h 2πfm(2n∆n+∆2 n)d2ξ¯0σψ cr¯0 i2 . (22) Therefore, we have |Ra(n, fm)| = σ 2 α · |Ia θ | · 1 r 1 + h 2πfm(2n∆… view at source ↗
Figure 3
Figure 3. Figure 3: The frequency correlation value. and |Is r | = 1 r 1 + h 2π∆mfsσψ c  1 − n2d2ξ¯0 r¯0 i2 . (32) Therefore, |Rs(∆m, n)| = σ 2 α · |Is θ | · 1 r 1 + h 2π∆mfsσψ c  1 − n2d2ξ¯0 r¯0 i2 . (33) Expression (33) shows that the inter-subcarrier correlation in near-field wideband channel depends jointly on the subcarrier spacing ∆m and the antenna index n. This space-frequency coupling originates from the beam spl… view at source ↗
Figure 4
Figure 4. Figure 4: Architecture of the proposed denoising network with a multi-scale attention mechanism. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Detailed structure of the proposed multi-scale attention block. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: NMSE performance comparison of different channel estimation [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: NMSE performance comparison of different channel estimation [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
read the original abstract

Extremely large-scale multiple-input multiple-output (XL-MIMO) is a key enabling technology for sixth-generation (6G) communication systems. Nevertheless, the increase in array aperture and signal bandwidth brings new challenges to wideband channel estimation in XL-MIMO systems. Motivated by recent advances in deep generative modeling, we propose a diffusion model-based method for near-field wideband channel estimation in XL-MIMO systems. We first analyze the statistical correlation of wideband channel and show that near-field wideband channel exhibits both spatial non-stationarity and beam split effects. Based on these observations, the channel estimation problem is formulated as a Bayesian posterior inference task, in which a diffusion model is employed to learn the prior distribution of the channel. To further enhance the representation of complex spatial-frequency channel structures, we design a denoising network with a multi-scale attention mechanism. In particular, the network extracts multi-scale spatial-frequency features via parallel convolutional branches with different receptive fields, and combines feature attention and spatial attention modules to adaptively emphasize critical channel features. This design enables more accurate modeling of near-field wideband channel distributions and consequently improves channel estimation performance. Experimental results demonstrate that the proposed method exhibits superior robustness to existing baseline schemes for XL-MIMO wideband channel estimation under different experimental settings.

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

3 major / 2 minor

Summary. The paper proposes a denoising diffusion model with a multi-scale attention denoising network for near-field wideband channel estimation in XL-MIMO systems. It first analyzes the statistical properties of the wideband channel to establish spatial non-stationarity and beam-split effects, formulates estimation as Bayesian posterior inference with the diffusion model supplying the prior, and designs a multi-scale convolutional attention network to capture complex spatial-frequency structures. Experimental results are reported to show superior robustness compared with existing baselines under varied settings.

Significance. If the performance claims hold under rigorous verification, the work would demonstrate a viable generative-modeling pathway for handling the high-dimensional, spatially non-stationary priors that appear in XL-MIMO wideband estimation. The explicit statistical analysis of non-stationarity and beam split, together with the tailored multi-scale attention architecture, supplies a concrete technical contribution to the emerging literature on diffusion models for wireless channel estimation.

major comments (3)
  1. [Abstract and §IV] Abstract and §IV: The central claim of superior robustness is stated without any quantitative details on training dataset size, exact baseline implementations, performance metrics (e.g., NMSE values), error bars, or ablation studies. This omission makes it impossible to judge whether reported gains are statistically meaningful or the result of favorable hyper-parameter tuning on simulated data.
  2. [§III] §III: The diffusion model is asserted to learn the true prior distribution that captures spatial non-stationarity and beam-split effects. However, no out-of-distribution tests or comparisons against channels generated under altered statistical assumptions are provided. Consequently, the performance advantage may simply reflect in-distribution matching rather than robust prior capture.
  3. [§IV] §IV: No controlled ablation isolating the multi-scale attention modules from a standard U-Net backbone is reported. Because the architectural innovation is presented as essential for modeling complex channel structures, the absence of such experiments weakens the attribution of gains to the proposed design.
minor comments (2)
  1. [Notation] Notation for the beam-split effect should be cross-referenced explicitly to the earlier correlation analysis to improve readability.
  2. [Figures] Figure captions should state the precise SNR range, array size, and bandwidth used for each curve.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments, which help clarify the presentation of our experimental results and the validation of the proposed approach. We address each major comment point by point below, indicating the specific revisions we will implement.

read point-by-point responses

  1. Referee: [Abstract and §IV] Abstract and §IV: The central claim of superior robustness is stated without any quantitative details on training dataset size, exact baseline implementations, performance metrics (e.g., NMSE values), error bars, or ablation studies. This omission makes it impossible to judge whether reported gains are statistically meaningful or the result of favorable hyper-parameter tuning on simulated data.

    Authors: We agree that the abstract and §IV would benefit from greater quantitative specificity to allow readers to assess the statistical significance of the results. While the manuscript reports comparative experiments under varied settings, we acknowledge that exact training dataset sizes, baseline implementation details, tabulated NMSE values, error bars, and ablations were not presented with sufficient precision. In the revised version we will expand the abstract with key numerical highlights and augment §IV with tables containing training sample counts, precise baseline descriptions, NMSE results accompanied by standard deviations across repeated trials, and additional ablation results. These changes will make the performance claims more transparent and reproducible. revision: yes

  2. Referee: [§III] §III: The diffusion model is asserted to learn the true prior distribution that captures spatial non-stationarity and beam-split effects. However, no out-of-distribution tests or comparisons against channels generated under altered statistical assumptions are provided. Consequently, the performance advantage may simply reflect in-distribution matching rather than robust prior capture.

    Authors: This observation correctly identifies a gap in validating the generality of the learned prior. The diffusion model is trained on channel realizations generated from the statistical model derived in the paper, which incorporates the analyzed spatial non-stationarity and beam-split effects. To demonstrate that the prior captures the underlying structure rather than merely fitting the training distribution, we will add out-of-distribution experiments in the revision. These will involve generating test channels under modified assumptions (for example, different array apertures or bandwidths that change the strength of non-stationarity and beam split) and reporting the resulting estimation performance. Such tests will provide direct evidence of robustness beyond exact in-distribution matching. revision: yes

  3. Referee: [§IV] §IV: No controlled ablation isolating the multi-scale attention modules from a standard U-Net backbone is reported. Because the architectural innovation is presented as essential for modeling complex channel structures, the absence of such experiments weakens the attribution of gains to the proposed design.

    Authors: We concur that a controlled ablation is required to isolate the contribution of the multi-scale attention components. The denoising network augments a convolutional backbone with parallel multi-scale branches and attention modules specifically to capture the spatial-frequency structure of near-field wideband channels. In the revised manuscript we will include an ablation study that replaces the proposed multi-scale attention modules with a standard U-Net backbone while keeping all other training and inference settings identical. NMSE results for both variants will be reported, thereby clarifying the performance gains attributable to the architectural choices. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the derivation chain

full rationale

The paper's chain consists of a statistical analysis of wideband channels (leading to observed non-stationarity and beam-split effects), a Bayesian posterior formulation, training a diffusion model with multi-scale attention to represent the prior, and empirical comparison against baselines. No quoted step equates a claimed result or prediction to its inputs by construction, self-citation, or renaming; the performance claims rest on experimental outcomes rather than definitional reduction. The method is self-contained with independent validation against external baseline schemes.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard assumptions of diffusion models and neural network expressivity plus domain assumptions about channel statistics; no new physical entities are invented.

free parameters (2)
  • Diffusion model parameters
    Neural network weights and diffusion schedule parameters are fitted to channel data to learn the prior.
  • Multi-scale attention hyperparameters
    Receptive field sizes, attention dimensions, and training hyperparameters chosen to fit the observed spatial-frequency structures.
axioms (2)
  • domain assumption Near-field wideband channels exhibit spatial non-stationarity and beam split effects
    Invoked in the abstract to motivate the problem formulation and network design.
  • domain assumption Diffusion models can learn the prior distribution of complex channel data
    Core modeling assumption for the Bayesian posterior inference task.

pith-pipeline@v0.9.0 · 5542 in / 1396 out tokens · 26315 ms · 2026-05-10T00:02:57.124886+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages · 1 internal anchor

  1. [1]

    Extremely large-scale MIMO: Fundamentals, challenges, solutions, and future directions,

    Z. Wang, J. Zhang, H. Du, W. E. I. Sha, B. Ai, D. Niyato, and M. Debbah, “Extremely large-scale MIMO: Fundamentals, challenges, solutions, and future directions,”IEEE Wireless Commun., vol. 31, no. 3, pp. 117–124, Jun. 2024

    work page 2024

  2. [2]

    Next generation advanced transceiver technologies for 6G and beyond,

    C. You, Y . Cai, Y . Liu, M. Di Renzo, T. M. Duman, A. Yener, and A. L. Swindlehurst, “Next generation advanced transceiver technologies for 6G and beyond,”IEEE J. Sel. Areas Commun., vol. 43, no. 3, pp. 582–627, Mar. 2025

    work page 2025

  3. [3]

    An introduction to millimeter-wave mobile broad- band systems,

    Z. Pi and F. Khan, “An introduction to millimeter-wave mobile broad- band systems,”IEEE Commun. Mag., vol. 49, no. 6, pp. 101–107, Jun. 2011

    work page 2011

  4. [4]

    Near-Field ISAC for THz Wireless Systems,

    F. Zhang, T. Mao, M. Li, M. Hua, J. Chen, C. Masouros, and Z. Wang, “Near-Field ISAC for THz Wireless Systems,”IEEE Netw., vol. 39, no. 6, pp. 54–61, 2025

    work page 2025

  5. [5]

    Near-field wideband beam training based on distance-dependent beam split,

    T. Zheng, M. Cui, Z. Wu, and L. Dai, “Near-field wideband beam training based on distance-dependent beam split,”IEEE Trans. Wireless Commun., vol. 24, no. 2, pp. 1278–1292, Feb. 2025

    work page 2025

  6. [6]

    Beam squint assisted user localization in near-field integrated sensing and communications systems,

    H. Luo, F. Gao, W. Yuan, and S. Zhang, “Beam squint assisted user localization in near-field integrated sensing and communications systems,”IEEE Trans. Wireless Commun., vol. 23, no. 5, pp. 4504– 4517, May 2024

    work page 2024

  7. [7]

    Near-field wideband beamforming for extremely large antenna arrays,

    M. Cui and L. Dai, “Near-field wideband beamforming for extremely large antenna arrays,”IEEE Trans. Wireless Commun., vol. 23, no. 10, pp. 13110–13124, Oct. 2024

    work page 2024

  8. [8]

    Near-field beam focusing for wireless power transfer with dynamic metasurface antennas,

    H. Zhang, N. Shlezinger, F. Guidi, A. Guerra, D. Dardari, M. F. Imani, and Y . C. Eldar, “Near-field beam focusing for wireless power transfer with dynamic metasurface antennas,”IEEE Internet Things J., vol. 12, no. 12, pp. 18 596–18 605, 2025

    work page 2025

  9. [9]

    Beam focusing for near-field multiuser MIMO communications,

    H. Zhang, N. Shlezinger, F. Guidi, D. Dardari, M. F. Imani, and Y . C. Eldar, “Beam focusing for near-field multiuser MIMO communications,” IEEE Trans. Wireless Commun., vol. 21, no. 9, pp. 7476–7490, Sep. 2022

    work page 2022

  10. [10]

    Structured OFDM modulation for XL-MIMO systems with dual-wideband effects,

    W. Huang, L. Xu, H. Zhang, C. Kai, C. Li, and Y . Huang, “Structured OFDM modulation for XL-MIMO systems with dual-wideband effects,” IEEE Trans. Wireless Commun., vol. 24, no. 2, pp. 1510–1525, Feb. 2025

    work page 2025

  11. [11]

    Knowledge and data dual-driven channel estimation and feedback for ultra-massive MIMO systems under hybrid-field beam squint effect,

    K. Wang, Z. Gao, S. Chen, B. Ning, G. Chen, Y . Su, Z. Wang, and H. V . Poor, “Knowledge and data dual-driven channel estimation and feedback for ultra-massive MIMO systems under hybrid-field beam squint effect,” IEEE Trans. Wireless Commun., vol. 23, no. 9, pp. 11 240–11 259, Sep. 2024

    work page 2024

  12. [12]

    UE position sensing in non-LoS propagation for U6G wideband XL-MIMO systems,

    Z. Lu, Y . Han, J. Huo, H. Xu, S. Jin, and M. Matthaiou, “UE position sensing in non-LoS propagation for U6G wideband XL-MIMO systems,” IEEE Commun. Lett., vol. 29, no. 11, pp. 2755–2759, Nov. 2025

    work page 2025

  13. [13]

    Near-field high-speed user sensing in wideband mmWave communications: Algorithms and bounds,

    H. Miao and M. Peng, “Near-field high-speed user sensing in wideband mmWave communications: Algorithms and bounds,”IEEE Trans. Signal Process., vol. 73, pp. 919–935, Jan. 2025

    work page 2025

  14. [14]

    Codebook design based on beam energy spread for extremely large- scale arrays,

    W. Huang, S. Jin, H. Zhang, F. Guidi, S. He, C. Kai, and Y . Huang, “Codebook design based on beam energy spread for extremely large- scale arrays,”IEEE Trans. Commun., vol. 73, no. 12, pp. 14 945–14 959, Dec. 2025

    work page 2025

  15. [15]

    Channel estimation for extremely large-scale MIMO: Far-field or near-field?

    M. Cui and L. Dai, “Channel estimation for extremely large-scale MIMO: Far-field or near-field?”IEEE Trans. Commun., vol. 70, no. 4, pp. 2663–2677, Apr. 2022

    work page 2022

  16. [16]

    Adaptive joint sparse Bayesian approaches for near-field channel estimation,

    Z. Zhu, R. Yang, C. Li, Y . Huang, and L. Yang, “Adaptive joint sparse Bayesian approaches for near-field channel estimation,”IEEE Trans. Wireless Commun., vol. 24, no. 3, pp. 2590–2605, Mar. 2025

    work page 2025

  17. [17]

    Non-stationary channel estimation for extremely large-scale MIMO,

    Y . Chen and L. Dai, “Non-stationary channel estimation for extremely large-scale MIMO,”IEEE Trans. Wireless Commun., vol. 23, no. 7, pp. 7683–7697, Jul. 2024

    work page 2024

  18. [18]

    Spatially non-stationary XL-MIMO channel estimation: A three-layer generalized approximate message passing method,

    A. Tang, J.-B. Wang, Y . Pan, W. Zhang, Y . Chen, H. Yu, and R. C. de Lamare, “Spatially non-stationary XL-MIMO channel estimation: A three-layer generalized approximate message passing method,”IEEE Trans. Signal Process., vol. 73, pp. 356–371, Dec. 2024

    work page 2024

  19. [19]

    Non-stationary near-field channel estimation for XL-MIMO systems with hybrid combining,

    X. Zhang and J. Zheng, “Non-stationary near-field channel estimation for XL-MIMO systems with hybrid combining,”IEEE Wireless Commun. Lett., vol. 13, no. 10, pp. 2727–2731, Oct. 2024

    work page 2024

  20. [20]

    Near-field channel estimation for extremely large-scale reconfigurable intelligent surface (XL-RIS)-aided wideband mmWave systems,

    S. Yang, C. Xie, W. Lyu, B. Ning, Z. Zhang, and C. Yuen, “Near-field channel estimation for extremely large-scale reconfigurable intelligent surface (XL-RIS)-aided wideband mmWave systems,”IEEE J. Sel. Areas Commun., vol. 42, no. 6, pp. 1567–1582, Jun. 2024

    work page 2024

  21. [21]

    Wideband channel estimation for THz massive MIMO,

    J. Tan and L. Dai, “Wideband channel estimation for THz massive MIMO,”China Commun., vol. 18, no. 5, pp. 66–80, May 2021

    work page 2021

  22. [22]

    MBPD: A robust algorithm for polar-domain channel estimation in near-field wideband XL-MIMO systems,

    H. Wang, P. Guo, X. Li, F. Wen, X. Wang, and A. Nallanathan, “MBPD: A robust algorithm for polar-domain channel estimation in near-field wideband XL-MIMO systems,”IEEE Internet Things J., vol. 12, no. 12, pp. 18 461–18 470, Jun. 2025

    work page 2025

  23. [23]

    Near-field wideband non-stationary channel estimation for XL-MIMO based on frequency-dependent visibility region,

    X. Wang, H. Chen, R. Jiang, T. Liu, J. Zhao, and Y . Xu, “Near-field wideband non-stationary channel estimation for XL-MIMO based on frequency-dependent visibility region,”IEEE Trans. Cogn. Commun. Netw., vol. 12, pp. 3987–4001, Jun. 2025

    work page 2025

  24. [24]

    Wideband near-field channel covariance estimation for XL-MIMO systems in the face of beam split,

    C. Ruan, Z. Zhang, H. Jiang, J. Dang, L. Wu, and H. Zhang, “Wideband near-field channel covariance estimation for XL-MIMO systems in the face of beam split,”IEEE Trans. Veh. Technol., vol. 74, no. 2, pp. 2912– 2926, Feb. 2025

    work page 2025

  25. [25]

    An adaptive and robust deep learning framework for THz ultra-massive MIMO channel estimation,

    W. Yu, Y . Shen, H. He, X. Yu, S. Song, J. Zhang, and K. B. Letaief, “An adaptive and robust deep learning framework for THz ultra-massive MIMO channel estimation,”IEEE J. Sel. Topics Signal Process., vol. 17, no. 4, pp. 761–776, Jun. 2023

    work page 2023

  26. [26]

    Deep learning-based near-field wideband channel estimation: A joint LISTA-CP approach,

    J. Yang, B. Ai, W. Chen, S. Yang, G. Shi, N. Wang, and C. Yuen, “Deep learning-based near-field wideband channel estimation: A joint LISTA-CP approach,”IEEE Trans. Veh. Technol., vol. 74, no. 9, pp. 14 041–14 053, Apr. 2025

    work page 2025

  27. [27]

    Deep CNN-based channel estimation for mmWave massive MIMO systems,

    P. Dong, H. Zhang, G. Y . Li, I. S. Gaspar, and N. NaderiAlizadeh, “Deep CNN-based channel estimation for mmWave massive MIMO systems,” IEEE J. Sel. Topics Signal Process., vol. 13, no. 5, pp. 989–1000, Sep. 2019

    work page 2019

  28. [28]

    NEFT: A unified transformer framework for efficient near-field CSI feedback in XL-MIMO systems,

    T. Mao, H. Li, S. Tan, P. Wang, G. Liu, R. Liu, L. Zhang, M. Hua, D. Zheng, Z. Wang, et al., “NEFT: A Unified Transformer Framework for Efficient Near-Field CSI Feedback in XL-MIMO Systems,”arXiv preprint arXiv:2509.12748, 2025

    work page arXiv 2025

  29. [29]

    Diffusion model based posterior sampling for noisy linear inverse problems.arXiv preprint arXiv:2211.12343,

    X. Meng and Y . Kabashima, “Diffusion model based posterior sampling for noisy linear inverse problems,”arXiv:2211.12343, 2022

    work page arXiv 2022

  30. [30]

    Pseudoinverse-guided diffusion models for inverse problems,

    J. Song, A. Vahdat, M. Mardani, and J. Kautz, “Pseudoinverse-guided diffusion models for inverse problems,” inProc. Int. Conf. Learn. Represent., 2023

    work page 2023

  31. [31]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” inProc. Int. Conf. Learn. Represent., 2021, 1-22

    work page 2021

  32. [32]

    On the asymptotic mean square error optimality of diffusion models,

    B. Fesl, B. B ¨ock, F. Strasser, M. Baur, M. Joham, and W. Utschick, “On the asymptotic mean square error optimality of diffusion models,” inProc. 28th Int. Conf. Artif. Intell. Statist., vol. 258. May 2025, pp. 505-513. 13

    work page 2025

  33. [33]

    A statistical model for indoor multipath propagation,

    A. Saleh and R. Valenzuela, “A statistical model for indoor multipath propagation,”IEEE J. Sel. Areas Commun., vol. 5, no. 2, pp. 128–137, Feb. 1987

    work page 1987

  34. [34]

    Simplified spatial correlation models for clustered MIMO channels with different array configura- tions,

    A. Forenza, D. J. Love, and R. W. Heath, “Simplified spatial correlation models for clustered MIMO channels with different array configura- tions,”IEEE Trans. Veh. Technol., vol. 56, no. 4, pp. 1924–1934, Jul. 2007

    work page 1924

  35. [35]

    A model for the multipath delay profile of fixed wireless channels,

    V . Erceg, D. Michelson, S. Ghassemzadeh, L. Greenstein, A. Rustako, P. Guerlain, M. Dennison, R. Roman, D. Barnickel, S. Wang, and R. Miller, “A model for the multipath delay profile of fixed wireless channels,” IEEE J. Sel. Areas Commun., vol. 17, no. 3, pp. 399–410, Mar. 1999

    work page 1999

  36. [36]

    A power-spectral theory of propagation in the mobile-radio environment,

    M. Gans, “A power-spectral theory of propagation in the mobile-radio environment,”IEEE Trans. Veh. Technol., vol. 21, no. 1, pp. 27–38, Feb. 1972

    work page 1972

  37. [37]

    Score-Based Generative Modeling through Stochastic Differential Equations

    Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Ermon, and B. Poole, “Score-based generative modeling through stochastic differential equations,”arXiv:2011.13456, 2020

    work page internal anchor Pith review Pith/arXiv arXiv 2011

  38. [38]

    A sparse-based channel estimation scheme for XL-RIS-assisted wireless communications with low pilot overheads,

    X. Bian, W. Xu, Y . Wang, C. Yuen, and Z. Cai, “A sparse-based channel estimation scheme for XL-RIS-assisted wireless communications with low pilot overheads,”IEEE Trans. Commun., vol. 74, pp. 367–380, Sep. 2025

    work page 2025

  39. [39]

    Channel estimation via orthogonal matching pursuit for hybrid MIMO systems in millimeter wave com- munications,

    J. Lee, G.-T. Gil, and Y . H. Lee, “Channel estimation via orthogonal matching pursuit for hybrid MIMO systems in millimeter wave com- munications,”IEEE Trans. Commun., vol. 64, no. 6, pp. 2370–2386, Apr. 2016

    work page 2016

  40. [40]

    Low-complexity on-grid channel estimation for partially-connected hybrid XL-MIMO,

    S. Kim and W. Choi, “Low-complexity on-grid channel estimation for partially-connected hybrid XL-MIMO,”IEEE Trans. Wireless Commun., vol. 25, pp. 3153–3169, Sep. 2025

    work page 2025

  41. [41]

    Near-field wideband channel estimation for ex- tremely large-scale MIMO,

    M. Cui and L. Dai, “Near-field wideband channel estimation for ex- tremely large-scale MIMO,”Sci. China Inf. Sci., vol. 66, no. 7, pp. 172303, Jun. 2023

    work page 2023

  42. [42]

    Cell- free distributed precoding without iterations on unreliable fronthaul by quadratic team learning,

    Z. Hong, J. Xue, X. Xia, T. Li, S. Xu, C. Li, D. Wang, and X. You, “Cell- free distributed precoding without iterations on unreliable fronthaul by quadratic team learning,”IEEE Trans. Commun., vol. 74, pp. 2047–2063, Dec. 2025

    work page 2047