pith. sign in

arxiv: 2605.26549 · v1 · pith:JZBS5GMXnew · submitted 2026-05-26 · 💻 cs.IT · eess.SP· math.IT

Joint Localization and Orientation with Triple-Beam Fingerprints in Massive MIMO-OFDM

Yu Zhao , Zhenzhou Jin , Jinke Tang , Li You , Chen Sun , Xiang-Gen Xia , Xiqi Gao This is my paper

Pith reviewed 2026-07-01 16:59 UTC · model grok-4.3

classification 💻 cs.IT eess.SPmath.IT
keywords triple-beam fingerprintlocalizationorientation estimationmassive MIMOOFDMTransformerDopplerfingerprinting
0
0 comments X

The pith

Triple-beam fingerprints incorporating Doppler enable joint position and motion direction estimation in massive MIMO-OFDM via a Transformer network.

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

The paper proposes triple-beam fingerprints that add Doppler information to capture both location and motion state, addressing the gap in existing fingerprints that only hold position data. It introduces LOA-Net, built around a mask-augmented detection Transformer for regression and a fusion-enhanced Transformer for direction classification, to process angle-delay and Doppler domain data while exploiting fingerprint sparsity. Simulations using 3GPP 38.901 indoor scenarios show the method yields higher localization accuracy than weighted K-nearest neighbors and both 2D and 3D convolutional networks, plus acceptable accuracy for motion direction. The authors first establish that TBF correlates with multipath components and that different TBFs are collinear, positioning it as a compact sparse fingerprint suitable for the network.

Core claim

TBF serves as an effective small-size sparse fingerprint because it correlates with multipath information and different TBFs exhibit collinearity; when fed to LOA-Net containing the MaskDETR-Reg module for position regression and the Fusion-TDC module for direction classification, the approach simultaneously estimates user position and motion direction with higher accuracy than WKNN or CNN baselines in 3GPP indoor simulations.

What carries the argument

The triple-beam fingerprint (TBF) that folds in Doppler information, shown via its multipath correlation and inter-TBF collinearity to act as a compact sparse representation, and the LOA-Net architecture that separates angle-delay processing from Doppler processing through dedicated Transformer modules.

If this is right

  • The method delivers significantly higher localization accuracy than WKNN, 2D CNNs, and 3D CNNs in the simulated indoor scenarios.
  • Motion direction estimation reaches satisfactory accuracy levels alongside the position estimates.
  • TBF is established as a compact sparse fingerprint through demonstrated correlation with multipath components and collinearity across TBF instances.

Where Pith is reading between the lines

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

  • If the collinearity property holds beyond the simulated indoor cases, TBF could reduce storage and matching costs in large-scale fingerprint databases.
  • Separating Doppler processing into its own Transformer branch may generalize to other wireless sensing tasks that require velocity alongside location.
  • The sparsity exploitation in LOA-Net suggests similar mask-augmented designs could improve efficiency when fingerprints are collected at lower sampling rates.

Load-bearing premise

That TBF qualifies as an effective small-size sparse fingerprint on the basis of its correlation with multipath information and the collinearity of different TBFs.

What would settle it

An experiment in the same 3GPP 38.901 indoor scenarios in which the proposed TBF-plus-LOA-Net method fails to exceed the localization accuracy of WKNN or the CNN baselines, or fails to produce satisfactory motion-direction estimates.

Figures

Figures reproduced from arXiv: 2605.26549 by Chen Sun, Jinke Tang, Li You, Xiang-Gen Xia, Xiqi Gao, Yu Zhao, Zhenzhou Jin.

Figure 1
Figure 1. Figure 1: Frame structure and massive MIMO-OFDM system [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The structure of TBF and its slices in each dimension. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The network architecture of the proposed LOA-Net. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The network architecture of the proposed MaskDETR-R [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The network architecture of the proposed Fusion-TDC [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Cumulative distribution functions of localization [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: Loss function values for different models. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: Confusion matrix for Fusion-TDC (5 km/h). [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
read the original abstract

With the widespread application of location-based services, fingerprint-based localization has demonstrated advantages in environments with complex signal propagation. Deep learning has significantly improved the efficiency of both offline training and online matching in localization processes. However, existing fingerprints only contain terminal position information without capturing motion states, and neural network designs have not fully incorporated structural features such as fingerprint sparsity. In this paper, we propose a triple-beam fingerprint (TBF) incorporating Doppler information and design a Transformer-based localization and orientation awareness network (LOA-Net) to simultaneously estimate user position and motion direction in massive multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems. We first show the correlation between TBF and multipath information, and investigate the collinearity of different TBFs, demonstrating that TBF is an effective small-size sparse fingerprint. Then, we propose LOA-Net containing a mask-augmented detection Transformer for regression (MaskDETR-Reg) module and a fusion-enhanced Transformer for direction classification (Fusion-TDC) module to process angle-delay domain information and Doppler domain information, respectively. Finally, in the simulation of indoor scenarios defined in 3GPP 38.901, the proposed method achieves significantly better localization accuracy than weighted $K$-nearest neighbors (WKNN), 2D and 3D convolutional neural networks (CNNs), and achieves satisfactory motion direction estimation accuracy.

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 triple-beam fingerprint (TBF) incorporating Doppler information as a small-size sparse fingerprint for joint user localization and motion direction estimation in massive MIMO-OFDM systems. It introduces LOA-Net, comprising a mask-augmented detection Transformer for regression (MaskDETR-Reg) module and a fusion-enhanced Transformer for direction classification (Fusion-TDC) module, to process angle-delay and Doppler domain data. Simulations in 3GPP 38.901 indoor scenarios are reported to yield significantly better localization accuracy than weighted K-nearest neighbors (WKNN) and 2D/3D CNN baselines, along with satisfactory motion direction estimation accuracy. The work first claims to demonstrate TBF's correlation with multipath information and collinearity across different TBFs.

Significance. If the TBF properties and performance gains hold under rigorous validation, the approach could advance fingerprint-based localization by incorporating motion-state information into sparse fingerprints and applying Transformer modules tailored to angle-delay and Doppler domains. The use of standardized 3GPP 38.901 scenarios provides a reproducible benchmark for comparison against conventional methods like WKNN and CNNs.

major comments (2)
  1. [TBF analysis section (preceding LOA-Net proposal)] The central claim that TBF constitutes an effective small-size sparse fingerprint rests on the asserted correlation with multipath information and collinearity of different TBFs (stated in the abstract as the first contribution). No equations defining the TBF construction, quantitative correlation metrics, collinearity measures, or supporting figures are referenced, which is load-bearing for attributing reported gains to the fingerprint rather than to the LOA-Net architecture or simulation setup.
  2. [Simulation results section] In the simulation results (3GPP 38.901 indoor scenarios), the claims of significantly better localization accuracy versus WKNN, 2D CNN, and 3D CNN lack reported error bars, number of Monte Carlo realizations, training/validation dataset sizes, or statistical tests. This undermines assessment of whether the gains are robust or attributable to TBF.
minor comments (2)
  1. [LOA-Net architecture description] Clarify the exact input dimensions and preprocessing steps for the angle-delay domain information fed to MaskDETR-Reg and the Doppler domain information fed to Fusion-TDC.
  2. Ensure all acronyms (TBF, LOA-Net, MaskDETR-Reg, Fusion-TDC) are defined at first use and used consistently.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment below and will revise the manuscript to incorporate the requested clarifications and additional details.

read point-by-point responses

  1. Referee: [TBF analysis section (preceding LOA-Net proposal)] The central claim that TBF constitutes an effective small-size sparse fingerprint rests on the asserted correlation with multipath information and collinearity of different TBFs (stated in the abstract as the first contribution). No equations defining the TBF construction, quantitative correlation metrics, collinearity measures, or supporting figures are referenced, which is load-bearing for attributing reported gains to the fingerprint rather than to the LOA-Net architecture or simulation setup.

    Authors: We agree that the TBF analysis requires explicit definitions and quantitative support to substantiate the claims. In the revised manuscript, we will add the mathematical construction of the TBF, quantitative metrics for its correlation with multipath components, measures of collinearity across TBFs (such as vector similarities), and corresponding figures. This will allow clearer attribution of performance improvements to the fingerprint properties. revision: yes

  2. Referee: [Simulation results section] In the simulation results (3GPP 38.901 indoor scenarios), the claims of significantly better localization accuracy versus WKNN, 2D CNN, and 3D CNN lack reported error bars, number of Monte Carlo realizations, training/validation dataset sizes, or statistical tests. This undermines assessment of whether the gains are robust or attributable to TBF.

    Authors: We acknowledge the need for greater statistical rigor in the results section. The revised version will specify the number of Monte Carlo realizations, training and validation dataset sizes, include error bars on performance plots, and report statistical significance tests comparing against the baselines to confirm robustness of the gains. revision: yes

Circularity Check

0 steps flagged

No circularity; derivation self-contained with independent analysis and external benchmarks

full rationale

The paper defines TBF, shows its correlation with multipath information and collinearity of different TBFs to establish it as an effective sparse fingerprint, then applies it in the LOA-Net architecture (MaskDETR-Reg and Fusion-TDC modules) for joint position and direction estimation. These steps rely on described signal processing properties and Transformer designs rather than reducing to fitted parameters renamed as predictions or self-citation chains. Results are validated against independent baselines (WKNN, 2D/3D CNNs) in 3GPP 38.901 indoor simulations, providing external falsifiability. No self-definitional loops, uniqueness theorems from prior author work, or ansatz smuggling via citation are present. The derivation chain is self-contained against the stated assumptions and benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 2 invented entities

Review based on abstract only; full paper unavailable so ledger is minimal. TBF and LOA-Net modules are presented as novel contributions without upstream derivation.

invented entities (2)
  • Triple-Beam Fingerprint (TBF) no independent evidence
    purpose: Fingerprint that incorporates Doppler information for motion state capture
    Introduced as new sparse fingerprint type correlated with multipath
  • LOA-Net with MaskDETR-Reg and Fusion-TDC modules no independent evidence
    purpose: Transformer-based network for joint localization and orientation estimation
    Proposed architecture to process angle-delay and Doppler domains

pith-pipeline@v0.9.1-grok · 5799 in / 1233 out tokens · 35290 ms · 2026-07-01T16:59:44.194285+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Channel Charting for Position and Orientation

    eess.SP 2026-06 unverdicted novelty 6.0

    Extends channel charting with an orientation triplet loss and alignment loss to estimate UE position and orientation from CSI, reaching accuracy close to supervised methods on real 5G NR measurements.

Reference graph

Works this paper leans on

52 extracted references · cited by 1 Pith paper

  1. [1]

    Joint localization and orientation assisted by massive MI MO-OFDM triple-beam fingerprints,

    Y . Zhao, Z. Jin, J. Tang, L. Y ou, C. Sun, X.-G. Xia, and X. Ga o, “Joint localization and orientation assisted by massive MI MO-OFDM triple-beam fingerprints,” in Proc. IEEE Wireless Commun. Netw. Conf. W orkshops (WCNC W orkshops), Kuala Lumpur, Malaysia, Apr. 2026, pp. 1–6

    2026

  2. [2]

    Integrated communications a nd localiza- tion for massive MIMO LEO satellite systems,

    L. Y ou, X. Qiang, Y . Zhu, F. Jiang, C. G. Tsinos, W. Wang, H. Wymeer- sch, X. Gao, and B. Ottersten, “Integrated communications a nd localiza- tion for massive MIMO LEO satellite systems,” IEEE Trans. Wireless Commun., vol. 23, no. 9, pp. 11 061–11 075, Sep. 2024

    2024

  3. [3]

    A survey of indoor localization systems and technologies,

    F. Zafari, A. Gkelias, and K. K. Leung, “A survey of indoor localization systems and technologies,” IEEE Commun. Surv. Tutorials , vol. 21, no. 3, pp. 2568–2599, 3rd Quart. 2019

    2019

  4. [4]

    Robust precoding for massive MIMO LEO satellite integrated commun ication and localization systems,

    Y . Zhu, L. Y ou, H. Zhou, Z. Jin, Q. Kong, and X. Gao, “Robust precoding for massive MIMO LEO satellite integrated commun ication and localization systems,” IEEE Commun. Lett. , vol. 29, no. 1, pp. 21– 25, Jan. 2025

    2025

  5. [5]

    Mult i- beam object-localization for millimeter-wave ISAC-aided connected autonomous vehicles,

    J. Singh, A. Gupta, A. K. Jagannatham, and L. Hanzo, “Mult i- beam object-localization for millimeter-wave ISAC-aided connected autonomous vehicles,” IEEE Trans. V eh. Technol. , vol. 74, no. 1, pp. 1725–1729, Jan. 2025

    2025

  6. [6]

    GDM4MMIMO: Generative diffusion models for massive MIMO communications,

    Z. Jin, L. Y ou, H. Zhou, Y . Wang, X. Liu, X. Gong, X. Gao, D. W . K. Ng, and X.-G. Xia, “GDM4MMIMO: Generative diffusion models for massive MIMO communications,” IEEE Commun. Mag. , vol. 64, no. 4, pp. 50–56, Apr. 2026

    2026

  7. [7]

    Near -field channel estimation for XL-MIMO: A deep generative model gui ded by side information,

    Z. Jin, L. Y ou, D. Wing Kwan Ng, X.-G. Xia, and X. Gao, “Near -field channel estimation for XL-MIMO: A deep generative model gui ded by side information,” IEEE Trans. Cognit. Commun. Networking , vol. 12, pp. 628–643, Dec. 2025

    2025

  8. [8]

    Convergent communication, sensing and localization in 6G systems: An overview of technologies, opportunities and challenges,

    C. De Lima, D. Belot, R. Berkvens, A. Bourdoux, D. Dardari , M. Guil- laud, M. Isomursu, E.-S. Lohan, Y . Miao, A. N. Barreto, M. R. K . Aziz, J. Saloranta, T. Sanguanpuak, H. Sarieddeen, G. Seco-Grana dos, J. Su- utala, T. Svensson, M. V alkama, B. V an Liempd, and H. Wymeers ch, “Convergent communication, sensing and localization in 6G systems: An overv...

    2021

  9. [9]

    S CNR maximization for MIMO ISAC assisted by fluid antenna system,

    Y . Y e, L. Y ou, H. Xu, A. Elzanaty, K.-K. Wong, and X. Gao, “S CNR maximization for MIMO ISAC assisted by fluid antenna system, ” IEEE Trans. V eh. Technol., vol. 74, no. 8, pp. 13 272–13 277, Aug. 2025

    2025

  10. [10]

    Radio positioning with EM proc essing of the spherical wavefront,

    F. Guidi and D. Dardari, “Radio positioning with EM proc essing of the spherical wavefront,” IEEE Trans. Wireless Commun., vol. 20, no. 6, pp. 3571–3586, Jun. 2021

    2021

  11. [11]

    Flui d antenna- assisted MIMO transmission exploiting statistical CSI,

    Y . Y e, L. Y ou, J. Wang, H. Xu, K.-K. Wong, and X. Gao, “Flui d antenna- assisted MIMO transmission exploiting statistical CSI,” IEEE Commun. Lett., vol. 28, no. 1, pp. 223–227, Jan. 2024

    2024

  12. [12]

    Distance-bas ed interpolation and extrapolation methods for RSS-based localization with indoor wire- less signals,

    J. Talvitie, M. Renfors, and E. S. Lohan, “Distance-bas ed interpolation and extrapolation methods for RSS-based localization with indoor wire- less signals,” IEEE Trans. V eh. Technol., vol. 64, no. 4, pp. 1340–1353, Apr. 2015

    2015

  13. [13]

    In tegrated statistical test of signal distributions and access point c ontributions for Wi-Fi indoor localization,

    M. Zhou, Y . Li, M. J. Tahir, X. Geng, Y . Wang, and W. He, “In tegrated statistical test of signal distributions and access point c ontributions for Wi-Fi indoor localization,” IEEE Trans. V eh. Technol. , vol. 70, no. 5, pp. 5057–5070, May 2021

    2021

  14. [14]

    A novel algorithm for multipath fingerprinting in indoor WLAN environments,

    S. H. Fang, T. N. Lin, and K. C. Lee, “A novel algorithm for multipath fingerprinting in indoor WLAN environments,” IEEE Trans. Wireless Commun., vol. 7, no. 9, pp. 3579–3588, Sep. 2008

    2008

  15. [15]

    A convex optimizat ion approach for NLOS error mitigation in TOA-based localizati on,

    H. Wu, L. Liang, X. Mei, and Y . Zhang, “A convex optimizat ion approach for NLOS error mitigation in TOA-based localizati on,” IEEE Signal Process Lett. , vol. 29, pp. 677–681, Jan. 2022

    2022

  16. [16]

    Optimum reference node deployment for TOA-based localization,

    K. Tong, X. Wang, A. Khabbazibasmenj, and A. Dounavis, “ Optimum reference node deployment for TOA-based localization,” in Proc. IEEE Int. Conf. Commun. (ICC) , London, United kingdom, 2015, pp. 3252– 3256

    2015

  17. [17]

    Toward n ear-ground localization: Modeling and applications for TOA ranging er ror,

    C. Xu, J. He, X. Zhang, P .-H. Tseng, and S. Duan, “Toward n ear-ground localization: Modeling and applications for TOA ranging er ror,” IEEE Trans. Antennas Propag. , vol. 65, no. 10, pp. 5658–5662, Oct. 2017

    2017

  18. [18]

    TDOA- based lo- calization for semi-static targets in NLOS environments,

    S. Li, M. Hedley, I. B. Collings, and D. Humphrey, “TDOA- based lo- calization for semi-static targets in NLOS environments,” IEEE Wireless Commun. Lett. , vol. 4, no. 5, pp. 513–516, Oct. 2015

    2015

  19. [19]

    Sensing user’s activity, channel, an d location with near-field extra-large-scale MIMO,

    L. Qiao, A. Liao, Z. Li, H. Wang, Z. Gao, X. Gao, Y . Su, P . Xi ao, L. Y ou, and D. W. K. Ng, “Sensing user’s activity, channel, an d location with near-field extra-large-scale MIMO,” IEEE Trans. Commun., vol. 72, no. 2, pp. 890–906, Feb. 2024

    2024

  20. [20]

    Unified near-field and far-field loca lization for AOA and hybrid AOA-TDOA positionings,

    Y . Wang and K. C. Ho, “Unified near-field and far-field loca lization for AOA and hybrid AOA-TDOA positionings,” IEEE Trans. Wireless Commun., vol. 17, no. 2, pp. 1242–1254, Feb. 2018

    2018

  21. [21]

    Exploiting AoA est imation accuracy for indoor localization: A weighted AoA-based app roach,

    Y . Zheng, M. Sheng, J. Liu, and J. Li, “Exploiting AoA est imation accuracy for indoor localization: A weighted AoA-based app roach,” IEEE Wireless Commun. Lett. , vol. 8, no. 1, pp. 65–68, Feb. 2019

    2019

  22. [22]

    Machine learning-bas ed finger- print positioning for massive MIMO systems,

    X. Gong, X. Y u, X. Liu, and X. Gao, “Machine learning-bas ed finger- print positioning for massive MIMO systems,” IEEE Access , vol. 10, pp. 89 320–89 330, Aug. 2022

    2022

  23. [23]

    Selection of signal sources influence at indoor positioning system,

    M. Luckner, S. Sowik, and P . Brida, “Selection of signal sources influence at indoor positioning system,” IEEE Trans. Wireless Commun., vol. 23, no. 1, pp. 45–57, Jan. 2024

    2024

  24. [24]

    Deep learning based fingerprint positioning for multi-cell mass ive MIMO- OFDM systems,

    X. Gong, A. Lu, X. Liu, X. Fu, X. Gao, and X.-G. Xia, “Deep learning based fingerprint positioning for multi-cell mass ive MIMO- OFDM systems,” IEEE Trans. V eh. Technol., vol. 73, no. 3, pp. 3832– 3849, Mar. 2024. 15

    2024

  25. [25]

    EnvCDiff: Joint re finement of environmental information and channel fingerprints via c onditional generative diffusion model,

    Z. Jin, L. Y ou, X.-G. Xia, and X. Gao, “EnvCDiff: Joint re finement of environmental information and channel fingerprints via c onditional generative diffusion model,” IEEE Trans. V eh. Technol., vol. 75, no. 4, pp. 6846–6851, Apr. 2025

    2025

  26. [26]

    Channe l acquisition for massive MIMO-OFDM with adjustable phase shift pilots,

    L. Y ou, X. Gao, A. L. Swindlehurst, and W. Zhong, “Channe l acquisition for massive MIMO-OFDM with adjustable phase shift pilots,” IEEE Trans. Signal Process. , vol. 64, no. 6, pp. 1461–1476, Mar. 2016

    2016

  27. [27]

    Sta- tistical CSI acquisition for multi-frequency massive MIMO systems,

    J. Tang, L. Y ou, X. Gong, C. Xie, X. Gao, X.-G. Xia, and X. S hi, “Sta- tistical CSI acquisition for multi-frequency massive MIMO systems,” IEEE Trans. Commun. , vol. 73, no. 11, pp. 11 798–11 813, Nov. 2025

    2025

  28. [28]

    Precoding design for joint synchronization and positioni ng in 5G integrated satellite communications,

    T. Chen, W. Wang, R. Ding, G. Seco-Granados, L. Y ou, and X . Gao, “Precoding design for joint synchronization and positioni ng in 5G integrated satellite communications,” in Proc. IEEE Global Commun. Conf. (GLOBECOM) , Madrid, Spain, 2021, pp. 01–06

    2021

  29. [29]

    Massive MIMO transmission for LEO satellite communications,

    L. Y ou, K.-X. Li, J. Wang, X. Gao, X.-G. Xia, and B. Otters ten, “Massive MIMO transmission for LEO satellite communications,” IEEE J. Sel. Areas Commun. , vol. 38, no. 8, pp. 1851–1865, Aug. 2020

    2020

  30. [30]

    Channel fingerprint construction for massive MIMO: A deep condition al genera- tive approach,

    Z. Jin, L. Y ou, X. Li, Z. Gao, Y . Liu, X.-G. Xia, and X. Gao, “Channel fingerprint construction for massive MIMO: A deep condition al genera- tive approach,” IEEE Trans. Wireless Commun., vol. 25, pp. 6096–6113, Dec. 2025

    2025

  31. [31]

    CF- CGN: Channel fingerprints extrapolation for multi-band massive MIMO trans- mission based on cycle-consistent generative networks,

    C. Xie, L. Y ou, Z. Jin, J. Tang, X. Gao, and X.-G. Xia, “CF- CGN: Channel fingerprints extrapolation for multi-band massive MIMO trans- mission based on cycle-consistent generative networks,” IEEE J. Sel. Areas Commun. , vol. 43, no. 11, pp. 3722–3736, Nov. 2025

    2025

  32. [32]

    Deterministic pilot design and channel estimation for dow nlink massive MIMO-OTFS systems in presence of the fractional doppler,

    D. Shi, W. Wang, L. Y ou, X. Song, Y . Hong, X. Gao, and G. Fet tweis, “Deterministic pilot design and channel estimation for dow nlink massive MIMO-OTFS systems in presence of the fractional doppler,” IEEE Trans. Wireless Commun. , vol. 20, no. 11, pp. 7151–7165, Nov. 2021

    2021

  33. [33]

    Massive MIMO-OFDM channel acquisition with t ime- frequency phase-shifted pilots,

    J. Tang, X. Gao, L. Y ou, D. Shi, J. Y ang, X.-G. Xia, X. Zhao , and P . Jiang, “Massive MIMO-OFDM channel acquisition with t ime- frequency phase-shifted pilots,” IEEE Trans. Commun. , vol. 73, no. 6, pp. 4520–4535, Jun. 2025

    2025

  34. [34]

    Channel acquisition for HF skywave massive MIMO-OFDM communicatio ns,

    D. Shi, L. Song, W. Zhou, X. Gao, C.-X. Wang, and G. Y e Li, “ Channel acquisition for HF skywave massive MIMO-OFDM communicatio ns,” IEEE Trans. Wireless Commun. , vol. 22, no. 6, pp. 4074–4089, Jun. 2023

    2023

  35. [35]

    CSI-tuples-ba sed 3D channel fingerprints construction assisted by multimoda l learning,

    C. Xie, L. Y ou, R. Chen, G. He, and X. Gao, “CSI-tuples-ba sed 3D channel fingerprints construction assisted by multimoda l learning,”

  36. [36]

    Available: https://arxiv.org/abs/2603

    [Online]. Available: https://arxiv.org/abs/2603. 25288

  37. [37]

    Single-site localiz ation based on a new type of fingerprint for massive MIMO-OFDM systems,

    X. Sun, X. Gao, G. Y . Li, and W. Han, “Single-site localiz ation based on a new type of fingerprint for massive MIMO-OFDM systems,” IEEE Trans. V eh. Technol., vol. 67, no. 7, pp. 6134–6145, Jul. 2018

    2018

  38. [38]

    Fingerprint-based lo calization for massive MIMO-OFDM system with deep convolutional neura l networks,

    X. Sun, C. Wu, X. Gao, and G. Y . Li, “Fingerprint-based lo calization for massive MIMO-OFDM system with deep convolutional neura l networks,” IEEE Trans. V eh. Technol., vol. 68, no. 11, pp. 10 846–10 857, Nov. 2019

    2019

  39. [39]

    Learn- ing to localize: A 3D CNN approach to user positioning in mass ive MIMO-OFDM systems,

    C. Wu, X. Yi, W. Wang, L. Y ou, Q. Huang, X. Gao, and Q. Liu, “ Learn- ing to localize: A 3D CNN approach to user positioning in mass ive MIMO-OFDM systems,” IEEE Trans. Wireless Commun., vol. 20, no. 7, pp. 4556–4570, Jul. 2021

    2021

  40. [40]

    Cooperative deep-lear ning positioning in mmWave 5G-advanced networks,

    B. C. Tedeschini and M. Nicoli, “Cooperative deep-lear ning positioning in mmWave 5G-advanced networks,” IEEE J. Sel. Areas Commun. , vol. 41, no. 12, pp. 3799–3815, Dec. 2023

    2023

  41. [41]

    CSI-fingerprinting in door local- ization via attention-augmented residual convolutional n eural network,

    B. Zhang, H. Sifaou, and G. Y . Li, “CSI-fingerprinting in door local- ization via attention-augmented residual convolutional n eural network,” IEEE Trans. Wireless Commun. , vol. 22, no. 8, pp. 5583–5597, Aug. 2023

    2023

  42. [42]

    A data-driven inertial navigation/bluetooth fusio n algorithm for indoor localization,

    J. Chen, B. Zhou, S. Bao, X. Liu, Z. Gu, L. Li, Y . Zhao, J. Zh u, and Q. Li, “A data-driven inertial navigation/bluetooth fusio n algorithm for indoor localization,” IEEE Sens. J. , vol. 22, no. 6, pp. 5288–5301, Mar. 2022

    2022

  43. [43]

    Deep-learning-based Wi- Fi indoor po- sitioning system using continuous CSI of trajectories,

    Z. Zhang, M. Lee, and S. Choi, “Deep-learning-based Wi- Fi indoor po- sitioning system using continuous CSI of trajectories,” Sensors, vol. 21, no. 17, p. 5776, Aug. 2021

    2021

  44. [44]

    Pilot reuse for massive MIMO transmission over spatially correlated rayleigh fadi ng channels,

    L. Y ou, X. Gao, X.-G. Xia, N. Ma, and Y . Peng, “Pilot reuse for massive MIMO transmission over spatially correlated rayleigh fadi ng channels,” IEEE Trans. Wireless Commun. , vol. 14, no. 6, pp. 3352–3366, Jun. 2015

    2015

  45. [45]

    Massive MIMO-OFDM channel acquisition with multi- group adjustable phase shift pilots,

    Y . Zhao, L. Y ou, J. Tang, M. Qian, B. Jiang, X.-G. Xia, and X. Gao, “Massive MIMO-OFDM channel acquisition with multi- group adjustable phase shift pilots,” IEEE Trans. Commun. , vol. 74, pp. 1702– 1716, Jan. 2026

    2026

  46. [46]

    G. H. Golub and C. F. V an Loan, Matrix Computations. Johns Hopkins University Press, 2013

    2013

  47. [47]

    SMART: Semantic-aware masked a ttention relational transformer for multi-label image recognition ,

    H. Wu, C. Xu, and H. Liu, “SMART: Semantic-aware masked a ttention relational transformer for multi-label image recognition ,” IEEE Signal Process Lett., vol. 29, pp. 2158–2162, Oct. 2022

    2022

  48. [48]

    Attention is all you need,

    A. V aswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jone s, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” in Proc. Adv. Neural Inf. Process. Syst. , vol. 30, 2017, pp. 1–11

    2017

  49. [49]

    End-to-end object detection with transform ers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillo v, and S. Zagoruyko, “End-to-end object detection with transform ers,” in Proc. Eur . Conf. Comput. Vis., 2020, pp. 213–229

    2020

  50. [50]

    QuaDRiGa: A 3- D multi-cell channel model with time evolution for enabling virtual field trials,

    S. Jaeckel, L. Raschkowski, K. B¨ orner, and L. Thiele, “ QuaDRiGa: A 3- D multi-cell channel model with time evolution for enabling virtual field trials,” IEEE Trans. Antennas Propag. , vol. 62, no. 6, pp. 3242–3256, Jun. 2014

    2014

  51. [51]

    2017, v12.9.0

    3GPP , “3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Evolved Universal Terrestrial Radio Access(E-UTRA); Physical Channels and Modulation (Releas e 12),” 3rd Generation Partnership Project (3GPP), Technical Spec ification (TS) 36.211, Mar. 2017, v12.9.0

    2017

  52. [52]

    3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Study on channel model for frequ encies from 0.5 to 100 GHz (Release 14),

    3GPP , “3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Study on channel model for frequ encies from 0.5 to 100 GHz (Release 14),” 3rd Generation Partnershi p Project (3GPP), Technical Report (TR) 38.901, Dec. 2017, v14.3.0

    2017