pith. machine review for the scientific record. sign in

arxiv: 2604.09017 · v1 · submitted 2026-04-10 · 💻 cs.NI

Recognition: unknown

Multimodal Large Language Model Enabled Robust Beamforming for HAP Downlink Communications

Xiaoyu Xing , Peng Yang , Guoquan Tao , Dingyi Lu , Zehui Xiong , Xianbin Cao
Authors on Pith no claims yet

Pith reviewed 2026-05-10 16:51 UTC · model grok-4.3

classification 💻 cs.NI
keywords high altitude platformbeamformingmultimodal LLMattitude forecastingproactive beam steeringdownlink communicationsQoS optimizationtelemetry data
0
0 comments X

The pith

A vision-language LLM trained on flight telemetry forecasts short-term high-altitude platform attitude changes to support proactive analog beamforming that raises user service ratio by 22.1 percent.

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

The paper aims to solve the problem of small attitude shifts in high-altitude platforms that misalign downlink beams and degrade communication quality. It shows that a multimodal vision-language model can extract patterns from multivariate flight data to predict those shifts over short horizons, even under shaking, and then use the predictions to update the analog beamformer ahead of time. An offline calibration step supplies error bounds that keep the proactive steering reliable, after which a lightweight QoS-aware optimizer admits users and allocates power while respecting instantaneous constraints. If the forecasts prove accurate enough in practice, the framework delivers measurable gains over standard reactive methods without requiring heavier online computation.

Core claim

The central claim is that a vision-language LLM learns from multivariate flight telemetry to forecast short-term HAP attitudes under platform shaking, an offline calibration procedure supplies reliable upper bounds on forecast error, and the resulting attitude predictions enable proactive analog beamformer updates together with a QoS-driven beamforming and admission algorithm that satisfies transmit-power and quality constraints, producing 22.1 percent higher user service ratio and 12.5 percent higher sum-rate than representative baselines while keeping total latency under 41 ms.

What carries the argument

The vision-language LLM (VL-LLM) that ingests multivariate flight telemetry for attitude forecasting, together with the offline forecast-error calibration that produces error upper bounds for proactive analog beam steering.

If this is right

  • The VL-LLM accurately captures attitude variations from telemetry, enabling delay-aware proactive steering.
  • The combined beamforming and admission method meets instantaneous power and QoS limits while outperforming baselines by the stated margins.
  • End-to-end latency remains low enough for practical deployment on real HAP systems.
  • The calibration step improves reliability of the forecasts without online retraining.

Where Pith is reading between the lines

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

  • Similar telemetry-driven forecasting could apply to attitude control on other mobile platforms such as UAVs or low-Earth-orbit satellites facing comparable mechanical disturbances.
  • Replacing the offline calibration with online uncertainty estimation might allow the system to adapt to changing flight conditions without separate training runs.
  • The approach indicates that LLMs can serve as lightweight predictors for physical dynamics in wireless systems, potentially reducing reliance on traditional dynamic models.

Load-bearing premise

Multivariate flight telemetry data contains enough information for the VL-LLM to predict short-term attitude changes under shaking with errors small enough that the calibrated bounds still allow reliable proactive beam alignment.

What would settle it

A controlled flight test in which measured attitude prediction errors exceed the calibrated bounds, causing beam misalignment that drops the achieved user service ratio below the levels of the reactive baselines.

Figures

Figures reproduced from arXiv: 2604.09017 by Dingyi Lu, Guoquan Tao, Peng Yang, Xianbin Cao, Xiaoyu Xing, Zehui Xiong.

Figure 1
Figure 1. Figure 1: Downlink beam misalignment for a HAP under wind [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Overall VL-LLM method for forecast-aided analog beamforming in [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Attitude trajectories and forecasting deviations for yaw, pitch, and roll. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗
read the original abstract

Small changes in high altitude platform (HAP) attitude can cause significant deviations in HAP downlink beam directions, thereby severely degrading HAP downlink communication performance. In this paper, we develop a multimodal large language model (LLM) enabled beamforming framework to achieve robust HAP downlink communications.Specifically, we design a vision-language LLM (VL-LLM) that learns from multivariate flight telemetry to forecast short-term HAP attitudes under platform shaking and support delay-aware proactive beam steering.We design an offline forecast-error calibration procedure to obtain upper bounds on forecast errors and improve the reliability of proactive analog beam steering.Based on the attitude forecasts, we proactively update the analog beamformer and propose a QoS-driven beamforming and admission method with a lightweight feasibility-enforcement step to satisfy instantaneous transmit-power and QoS requirements.Simulation results indicate that the designed VL-LLM can accurately capture changes in the HAP attitude and the proposed beamforming method achieves a 22.1% higher user service ratio and a 12.5% higher sum-rate than representative baselines.The measured mean and p99 total latencies are 36.24 ms and 40.13 ms, respectively, supporting practical delay-aware deployment.

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 develops a vision-language large language model (VL-LLM) framework for robust beamforming in high-altitude platform (HAP) downlink communications. It trains the VL-LLM on multivariate flight telemetry to forecast short-term HAP attitudes under platform shaking, performs an offline forecast-error calibration to derive upper bounds, and uses the forecasts for proactive analog beamforming combined with a QoS-driven admission control that includes a lightweight feasibility-enforcement step to meet instantaneous power and QoS constraints. Simulation results claim that the VL-LLM accurately captures attitude changes and that the method achieves a 22.1% higher user service ratio and 12.5% higher sum-rate than representative baselines, with mean and p99 total latencies of 36.24 ms and 40.13 ms supporting practical deployment.

Significance. If the simulation results generalize, the work addresses a real engineering challenge in HAP systems where attitude perturbations cause beam misalignment. The combination of multimodal LLM-based forecasting, offline error calibration, and proactive QoS-aware beamforming with feasibility enforcement is a novel application of large models to wireless physical-layer problems. The reported latency figures are a positive indicator for delay-sensitive use. However, the significance is limited by the closed-loop simulation setting and the absence of evidence that the telemetry distribution or shaking model matches real HAP flight data.

major comments (2)
  1. [Abstract] Abstract: The headline performance claims (22.1% higher user service ratio and 12.5% higher sum-rate) are obtained by feeding VL-LLM attitude forecasts into proactive analog beamforming plus QoS-driven admission. These gains are load-bearing for the central contribution, yet the abstract provides no quantitative forecast accuracy metrics (e.g., attitude prediction MSE or comparison against Kalman-filter or ARIMA baselines), no description of the telemetry data generation process, and no statistical significance tests or number of Monte-Carlo runs. Without these, it is impossible to verify that the reported improvements are robust rather than artifacts of the simulation setup.
  2. [Abstract] Abstract: The offline forecast-error calibration procedure is presented as the mechanism that yields reliable upper bounds for proactive steering. The central claim that the method satisfies instantaneous power and QoS constraints therefore rests on the validity of these bounds. The abstract gives no formulation of how the bounds are computed from the error distribution, no validation that the multivariate telemetry and shaking model used for training/validation match real HAP dynamics, and no sensitivity analysis showing what happens when the real error tails are heavier than assumed. This is a load-bearing assumption for the feasibility-enforcement step.
minor comments (2)
  1. [Abstract] The abstract states that the VL-LLM 'can accurately capture changes in the HAP attitude' but does not report any concrete accuracy figures or ablation studies; adding these (even in a table) would strengthen the presentation.
  2. [Abstract] The latency numbers (mean 36.24 ms, p99 40.13 ms) are useful, but the manuscript should clarify whether they include VL-LLM inference time or only the beamforming and admission computation.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive comments, which help clarify the presentation of our contributions. We address each major comment below, agreeing where revisions are needed to improve transparency and robustness of the claims.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The headline performance claims (22.1% higher user service ratio and 12.5% higher sum-rate) are obtained by feeding VL-LLM attitude forecasts into proactive analog beamforming plus QoS-driven admission. These gains are load-bearing for the central contribution, yet the abstract provides no quantitative forecast accuracy metrics (e.g., attitude prediction MSE or comparison against Kalman-filter or ARIMA baselines), no description of the telemetry data generation process, and no statistical significance tests or number of Monte-Carlo runs. Without these, it is impossible to verify that the reported improvements are robust rather than artifacts of the simulation setup.

    Authors: We agree that the abstract should be strengthened with supporting quantitative details. The full manuscript (Section IV-B) reports attitude prediction MSE for the VL-LLM versus Kalman filter and ARIMA baselines on the validation set, along with the telemetry generation process in Section III-A (multivariate flight model with shaking perturbations drawn from standard HAP dynamics). All results are averaged over 1000 independent Monte-Carlo runs with reported standard deviations. We will revise the abstract to concisely include forecast accuracy metrics and the simulation run count, ensuring the performance claims are better contextualized. revision: yes

  2. Referee: [Abstract] Abstract: The offline forecast-error calibration procedure is presented as the mechanism that yields reliable upper bounds for proactive steering. The central claim that the method satisfies instantaneous power and QoS constraints therefore rests on the validity of these bounds. The abstract gives no formulation of how the bounds are computed from the error distribution, no validation that the multivariate telemetry and shaking model used for training/validation match real HAP dynamics, and no sensitivity analysis showing what happens when the real error tails are heavier than assumed. This is a load-bearing assumption for the feasibility-enforcement step.

    Authors: The calibration procedure is formulated in Section III-C: upper bounds are computed as the 95th percentile of the empirical forecast-error norm distribution obtained from the held-out validation telemetry. We will add a brief statement of this formulation to the abstract. A sensitivity analysis under heavier-tailed errors will be added to the revised simulation results. However, the work is simulation-based and does not include direct validation against real proprietary HAP flight data; the shaking model follows published HAP dynamics literature. revision: partial

standing simulated objections not resolved
  • Direct validation of the telemetry distribution and shaking model against real high-altitude platform flight data (the study uses a simulated environment based on standard models, and real proprietary telemetry is unavailable).

Circularity Check

0 steps flagged

No circularity detected; derivation relies on independent training, calibration, and simulation validation

full rationale

The paper's chain consists of training a VL-LLM on multivariate telemetry to produce attitude forecasts, running a separate offline calibration to bound forecast errors, feeding those forecasts into a QoS-driven analog beamforming optimization with feasibility enforcement, and measuring service ratio and sum-rate gains against external baselines in simulation. No step reduces by construction to its inputs: the forecasts are learned outputs, not tautological definitions; the calibration is an independent procedure; and the reported 22.1% and 12.5% improvements are empirical simulation results, not renamed fitted quantities or self-citation chains. The derivation is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Based on abstract only; the central claims rest on domain assumptions about data sufficiency for prediction and calibration reliability rather than new mathematical derivations.

axioms (2)
  • domain assumption Multivariate flight telemetry contains sufficient predictive information for short-term HAP attitude changes
    Invoked when designing the VL-LLM to forecast attitudes from sensor data.
  • domain assumption Offline calibration can produce usable upper bounds on forecast errors for proactive steering
    Used to justify reliability of the beamforming updates.

pith-pipeline@v0.9.0 · 5517 in / 1375 out tokens · 56277 ms · 2026-05-10T16:51:24.118647+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

59 extracted references · 8 canonical work pages

  1. [1]

    Survey on near-space information networks: Channel modeling, transmission, and networking perspectives,

    X. Cao, X. Su, P. Yang, Y . Gao, D. Oliver Wu, and T. Q. S. Quek, “Survey on near-space information networks: Channel modeling, transmission, and networking perspectives,”IEEE Commun. Surveys Tuts., vol. 28, pp. 672–715, 2026

    2026

  2. [2]

    Combined control of coverage area and HAPS deployment in hybrid FSO/RF SAGIN,

    K. Mashiko, Y . Kawamoto, N. Kato, K. Yoshida, and M. Ariyoshi, “Combined control of coverage area and HAPS deployment in hybrid FSO/RF SAGIN,”IEEE Trans. Veh. Technol., vol. 74, no. 7, pp. 10 819– 10 828, 2025

    2025

  3. [3]

    NOMA- oriented spectrum sensing for joint HAP and HEO nonterrestrial uplink communications,

    T. Xu, J. Zhu, Y . Xu, P. Peng, X. Chen, Q. Wu, and D. Niyato, “NOMA- oriented spectrum sensing for joint HAP and HEO nonterrestrial uplink communications,”IEEE Internet Things J., vol. 12, no. 16, pp. 34 603– 34 606, 2025

    2025

  4. [4]

    Aerial ISAC: A HAPS-assisted integrated sensing, communications and computing framework for enhanced coverage and security,

    A. M. Benaya, M. S. Hassan, M. H. Ismail, and T. Landolsi, “Aerial ISAC: A HAPS-assisted integrated sensing, communications and computing framework for enhanced coverage and security,”IEEE Trans. Green Commun. Networking, vol. 9, no. 4, pp. 2101–2114, 2025. 15

    2025

  5. [5]

    Mobility- aware multicast orchestration for low-altitude UA Vs with integrated terrestrial and non-terrestrial networks,

    F. Wang, H. Yang, S. Zhang, J. Shi, Z. Li, and T. Q. S. Quek, “Mobility- aware multicast orchestration for low-altitude UA Vs with integrated terrestrial and non-terrestrial networks,”IEEE Trans. Commun., vol. 73, no. 11, pp. 12 854–12 868, 2025

    2025

  6. [6]

    Survey on HA2G and A2A channel measurement and modeling for airborne maneuvering networks,

    P. An, S. Yang, J. Sun, J. Li, P. Yang, and X. Cao, “Survey on HA2G and A2A channel measurement and modeling for airborne maneuvering networks,”Chin. J. Aeronaut., p. 104036, early access, Dec. 24, 2025, DOI:10.1016/j.cja.2025.104036

    work page doi:10.1016/j.cja.2025.104036 2025

  7. [7]

    Optimizing network performance and resource allocation in HAPS-UA V integrated sensing and communication systems for 6G,

    P. Kanani, M. Javad Omidi, M. Modarres-Hashemi, and H. Yanikomeroglu, “Optimizing network performance and resource allocation in HAPS-UA V integrated sensing and communication systems for 6G,”IEEE Trans. Wireless Commun., vol. 25, pp. 4098–4112, 2026

    2026

  8. [8]

    Sustainable UA V mobility support in integrated terrestrial and non-terrestrial networks,

    F. Wang, S. Zhang, J. Shi, Z. Li, and T. Q. S. Quek, “Sustainable UA V mobility support in integrated terrestrial and non-terrestrial networks,” IEEE Trans. Wireless Commun., vol. 23, no. 11, pp. 17 115–17 128, 2024

    2024

  9. [9]

    Deep joint semantic coding and beamforming for near-space airship- borne massive MIMO network,

    M. Wu, Z. Gao, Z. Wang, D. Niyato, G. K. Karagiannidis, and S. Chen, “Deep joint semantic coding and beamforming for near-space airship- borne massive MIMO network,”IEEE J. Sel. Areas Commun., vol. 43, no. 1, pp. 260–278, Jan. 2025

    2025

  10. [10]

    Spatially sparse precoding in millimeter wave MIMO systems,

    O. E. Ayach, S. Rajagopal, S. Abu-Surra, Z. Pi, and R. W. Heath, “Spatially sparse precoding in millimeter wave MIMO systems,”IEEE Trans. Wireless Commun., vol. 13, no. 3, pp. 1499–1513, 2014

    2014

  11. [11]

    Hybrid digital and analog beamforming design for large-scale antenna arrays,

    F. Sohrabi and W. Yu, “Hybrid digital and analog beamforming design for large-scale antenna arrays,”IEEE J. Sel. Top. Signal Process., vol. 10, no. 3, pp. 501–513, 2016

    2016

  12. [12]

    A generalized pointing error model for FSO links with fixed-wing UA Vs for 6G: Analysis and trajectory optimization,

    H.-J. Moon, C.-B. Chae, K.-K. Wong, and M.-S. Alouini, “A generalized pointing error model for FSO links with fixed-wing UA Vs for 6G: Analysis and trajectory optimization,”IEEE Trans. Wireless Commun., vol. 24, no. 7, pp. 5723–5737, 2025

    2025

  13. [13]

    Robust beamforming for UA V communication with jittering effects,

    K. Lee, D. You, H. Noh, and C. Lee, “Robust beamforming for UA V communication with jittering effects,”IEEE Wireless Commun. Lett., vol. 14, no. 1, pp. 48–52, 2025

    2025

  14. [14]

    A survey of beam management for mmwave and THz communications towards 6G,

    Q. Xue, C. Ji, S. Ma, J. Guo, Y . Xu, Q. Chen, and W. Zhang, “A survey of beam management for mmwave and THz communications towards 6G,”IEEE Commun. Surv. Tutorials, vol. 26, no. 3, 2024

    2024

  15. [15]

    When future communications shift toward narrow beams: A forward looking survey on pointing errors and alignment limits,

    M. Ghanbari, M. T. Dabiri, O. S. Badarneh, M. Hasna, Y . H. Al-Badarneh, M. K. Alshawaqfeh, and K. Qaraqe, “When future communications shift toward narrow beams: A forward looking survey on pointing errors and alignment limits,” 2025,arXiv:2512.20874

    work page arXiv 2025

  16. [16]

    Reliable beam tracking on high-altitude platform for millimeter wave high-speed railway,

    A. P. Salmeno and I. Zakia, “Reliable beam tracking on high-altitude platform for millimeter wave high-speed railway,”IEEE Access, vol. 12, pp. 71 997–72 012, 2024

    2024

  17. [17]

    GPS-aided deep learning for beam prediction and tracking in UA V mmwave communication,

    V . Ardianto Nugroho and B. Moo Lee, “GPS-aided deep learning for beam prediction and tracking in UA V mmwave communication,”IEEE Access, vol. 13, pp. 117 065–117 077, 2025

    2025

  18. [18]

    Combating beam misalignment in mmwave UA V networks: An attitude compensation- based method,

    Y . Ke, W. Xu, W. Ni, X. Yuan, and D. Niyato, “Combating beam misalignment in mmwave UA V networks: An attitude compensation- based method,”IEEE Trans. Veh. Technol., vol. 74, no. 2, pp. 3521–3526, 2025

    2025

  19. [19]

    WirelessLLM: Empowering large language models towards wireless intelligence,

    J. Shao, J. Tong, Q. Wu, W. Guo, Z. Li, Z. Lin, and J. Zhang, “WirelessLLM: Empowering large language models towards wireless intelligence,”J. Commun. Inf. Networks, vol. 9, no. 2, pp. 99–112, 2024

    2024

  20. [20]

    TSpec-LLM: An open- source dataset for LLM understanding of 3GPP specifications,

    R. Nikbakht, M. Benzaghta, and G. Geraci, “TSpec-LLM: An open- source dataset for LLM understanding of 3GPP specifications,” 2024, arXiv:2406.01768

    work page arXiv 2024

  21. [21]

    Telecom language models: Must they be large?

    N. Piovesan, A. D. Domenico, and F. Ayed, “Telecom language models: Must they be large?” 2024,arXiv:2403.04666

    work page arXiv 2024

  22. [22]

    Telco-rag: Navigating the challenges of retrieval-augmented language models for telecommunications,

    A.-L. Bornea, F. Ayed, A. D. Domenico, N. Piovesan, and A. Maatouk, “Telco-RAG: Navigating the challenges of retrieval-augmented language models for telecommunications,” 2024,arXiv:2404.15939

    work page arXiv 2024

  23. [23]

    Generative AI agents with large language model for satellite networks via a mixture of experts transmission,

    R. Zhang, H. Du, Y . Liu, D. Niyato, J. Kang, Z. Xiong, A. Jamalipour, and D. In Kim, “Generative AI agents with large language model for satellite networks via a mixture of experts transmission,”IEEE J. Sel. Areas Commun., vol. 42, no. 12, pp. 3581–3596, 2024

    2024

  24. [24]

    Large models for aerial edges: An edge-cloud model evolution and communication paradigm,

    S. Zhang, Q. Liu, K. Chen, B. Di, H. Zhang, W. Yang, D. Niyato, Z. Han, and H. V . Poor, “Large models for aerial edges: An edge-cloud model evolution and communication paradigm,”IEEE J. Sel. Areas Commun., vol. 43, no. 1, pp. 21–35, 2025

    2025

  25. [25]

    APEG: Adaptive physical layer authentication with channel extrapolation and generative AI,

    X. Cheng, R. Meng, X. Xu, H. Gao, P. Zhang, and D. Niyato, “APEG: Adaptive physical layer authentication with channel extrapolation and generative AI,”IEEE Trans. Inf. Forensics Secur., vol. 21, pp. 1257–1272, 2026

    2026

  26. [26]

    WMMSE-based joint transceiver design for multi-RIS-assisted cell-free networks using hybrid CSI,

    X. Pan, Z. Zheng, X. Huang, and Z. Fei, “WMMSE-based joint transceiver design for multi-RIS-assisted cell-free networks using hybrid CSI,”IEEE Trans. Wireless Commun., vol. 24, no. 9, pp. 7654–7669, 2025

    2025

  27. [27]

    Model-driven distributed WMMSE for downlink massive MIMO systems,

    N. Zhou, Z. Wang, Q. Shi, W. Xu, and Y . Huang, “Model-driven distributed WMMSE for downlink massive MIMO systems,”IEEE Wireless Commun. Lett., vol. 14, no. 11, pp. 3675–3679, 2025

    2025

  28. [28]

    Enhancing WMMSE with EVM feedback through bayesian optimization,

    Z. Zhao, X. Guan, F. Xu, and Q. Shi, “Enhancing WMMSE with EVM feedback through bayesian optimization,”IEEE Trans. Veh. Technol., vol. 74, no. 10, pp. 16 579–16 583, 2025

    2025

  29. [29]

    Joint robust beamforming design for RIS-assisted ISAC systems with hardware impairments,

    Z. Xinyu, Z. Lijuan, L. Shidang, Z. Yanping, C. Jianbin, and L. Chunguo, “Joint robust beamforming design for RIS-assisted ISAC systems with hardware impairments,”AEU Int. J. Electron. Commun., vol. 206, p. 156209, 2026

    2026

  30. [30]

    Joint optimization of multi-beam configuration and resource allocation for low-earth orbit satellites,

    Z. Bowen, Z. Pan, P. Zihao, Z. Wenming, Y . Ning, and G. Daoxing, “Joint optimization of multi-beam configuration and resource allocation for low-earth orbit satellites,”Phys. Commun., vol. 75, p. 103001, 2026

    2026

  31. [31]

    Distilling knowledge from resource management algorithms to neural networks: A unified training assistance approach,

    L. Ma, N. Cheng, X. Wang, Z. Yin, H. Zhou, and W. Quan, “Distilling knowledge from resource management algorithms to neural networks: A unified training assistance approach,” inProc. IEEE 98th Veh. Technol. Conf. (VTC2023-Fall), Dec. 2023, pp. 1–5

    2023

  32. [32]

    Beamforming for pinching antenna systems (PASS): KKT-guided large model learning,

    X. Xu, Y . Liu, X. Mu, D. Gan, and A. Nallanathan, “Beamforming for pinching antenna systems (PASS): KKT-guided large model learning,” inProc. IEEE/CIC Int. Conf. Commun. China (ICCC), Aug. 2025, pp. 1–6

    2025

  33. [33]

    Differentiable projection-based learn to optimize in wireless network-part I: Convex constrained (non-)convex programming,

    X. Wang, X. Zhao, and N. Cheng, “Differentiable projection-based learn to optimize in wireless network-part I: Convex constrained (non-)convex programming,” 2025,arXiv:2502.00053

    work page arXiv 2025

  34. [34]

    HoP: Homeomorphic polar learning for hard constrained optimization,

    K. Deng, H. Zhang, J. Lu, and H. Sun, “HoP: Homeomorphic polar learning for hard constrained optimization,” 2025,arXiv:2502.00304

    work page arXiv 2025

  35. [35]

    Unitime: A language-empowered unified model for cross-domain time series forecasting,

    X. Liu, J. Hu, Y . Li, S. Diao, Y . Liang, B. Hooi, and R. Zimmermann, “Unitime: A language-empowered unified model for cross-domain time series forecasting,”CoRR, vol. abs/2310.09751, 2023

    work page arXiv 2023

  36. [36]

    Time-LLM: Time series forecasting by reprogramming large language models,

    M. Jin, S. Wang, L. Ma, Z. Chu, J. Y . Zhang, X. Shi, P.-Y . Chen, Y . Liang, Y .-F. Li, S. Pan, and Q. Wen, “Time-LLM: Time series forecasting by reprogramming large language models,” inProc. 12th Int. Conf. Learn. Represent. (ICLR), May 2024, pp. 1–24

    2024

  37. [37]

    Beam prediction based on large language models,

    Y . Sheng, K. Huang, L. Liang, P. Liu, S. Jin, and G. Ye Li, “Beam prediction based on large language models,”IEEE Wireless Commun. Lett., vol. 14, no. 5, pp. 1406–1410, 2025

    2025

  38. [38]

    UniTime: A language-empowered unified model for cross-domain time series forecasting,

    X. Liu, J. Hu, Y . Li, S. Diao, Y . Liang, B. Hooi, and R. Zimmermann, “UniTime: A language-empowered unified model for cross-domain time series forecasting,” inProc. ACM Web Conf.Association for Computing Machinery, May 2024, p. 4095–4106

    2024

  39. [39]

    Timemixer++: A general time series pattern machine for universal predictive analysis,

    S. Wang, J. LI, X. Shi, Z. Ye, B. Mo, W. Lin, J. Shengtong, Z. Chu, and M. Jin, “Timemixer++: A general time series pattern machine for universal predictive analysis,” inProc. 13th Int. Conf. Learn. Represent. (ICLR), Apr. 2025, pp. 1–37

    2025

  40. [40]

    Timesnet: Temporal 2d-variation modeling for general time series analysis,

    H. Wu, T. Hu, Y . Liu, H. Zhou, J. Wang, and M. Long, “Timesnet: Temporal 2d-variation modeling for general time series analysis,” in Proc. 11th Int. Conf. Learn. Represent. (ICLR), Apr. 2023, pp. 1–23

    2023

  41. [41]

    An image is worth 16×16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikovet al., “An image is worth 16×16 words: Transformers for image recognition at scale,” inProc. 9th Int. Conf. Learn. Represent. (ICLR), 2021

    2021

  42. [42]

    CuIoT: Advancing Network Connectivity with Motif Knowledge-Centric for Robust Topology ,

    N. Chen, S. Zhang, X. Zhou, X. Huang, D. O. Wu, and T. Qiu, “ CuIoT: Advancing Network Connectivity with Motif Knowledge-Centric for Robust Topology ,”IEEE Trans. Knowl. Data Eng., no. 01, pp. 1–14, Mar. 5555

  43. [43]

    LEGO- Motif: Enhancing IoT topology robustness with evolutionary motif-based generation,

    N. Chen, T. Qiu, X. Zhou, S. Zhang, W. Si, and X. Wang, “LEGO- Motif: Enhancing IoT topology robustness with evolutionary motif-based generation,”IEEE Trans. Syst. Man Cybern.: Syst., pp. 1–14, 2026

    2026

  44. [44]

    Fast robustness enhancement for dynamic IIoT topology with adaptive bayesian learning,

    N. Chen, S. Zhang, X. Zhou, S. Cao, and T. Qiu, “Fast robustness enhancement for dynamic IIoT topology with adaptive bayesian learning,” IEEE Trans. Mob. Comput., vol. 24, no. 10, pp. 10 886–10 899, 2025

    2025

  45. [45]

    A distributed co-evolutionary optimization method with motif for large- scale IoT robustness,

    N. Chen, T. Qiu, X. Zhou, S. Zhang, W. Si, and D. Oliver Wu, “A distributed co-evolutionary optimization method with motif for large- scale IoT robustness,”IEEE/ACM Trans. Networking, vol. 32, no. 5, pp. 4085–4098, 2024

    2024

  46. [46]

    Robust beam control for terahertz drone networks,

    M. Khemiri, M. Erdem, A. Saeed, O. Gurbuz, and M.-S. Alouini, “Robust beam control for terahertz drone networks,”IEEE Trans. Aerosp. Electron. Syst., vol. 61, no. 5, 2025

    2025

  47. [47]

    Learning with 3D rotations, a hitchhiker’s guide to SO(3),

    A. R. Geist, J. Frey, M. Zhobro, A. Levina, and G. Martius, “Learning with 3D rotations, a hitchhiker’s guide to SO(3),” inProc. 41th Int. Conf. Machine Learning (ICML), vol. 235, Jul 2024, pp. 15 331–15 350

    2024

  48. [48]

    A general framework for multi- step ahead adaptive conformal heteroscedastic time series forecasting,

    M. Sousa, A. M. Tom ´e, and J. Moreira, “A general framework for multi- step ahead adaptive conformal heteroscedastic time series forecasting,” Neurocomputing, vol. 608, p. 128434, 2024

    2024

  49. [49]

    Adaptive hybrid beamforming for UA V mmwave communications against asymmetric jitter,

    W. Chen, C. Liu, W. Wang, M. Peng, and W. Zhang, “Adaptive hybrid beamforming for UA V mmwave communications against asymmetric jitter,”IEEE Trans. Wireless Commun., vol. 23, no. 8, pp. 9432–9445, Feb. 2024. 16

    2024

  50. [50]

    SINR maximizing distributionally robust adaptive beamforming,

    K. H. Irani, Y . Huang, and S. A. V orobyov, “SINR maximizing distributionally robust adaptive beamforming,”IEEE Trans. Signal Process., vol. 73, pp. 2542–2557, 2025

    2025

  51. [51]

    Radiorchestra: Proactive management of millimeter-wave self-backhauled small cells via joint optimization of beamforming, user association, rate selection, and admission control,

    L. F. Abanto-Leon, A. Asadi, A. Garcia-Saavedra, G. H. Sim, and M. Hollick, “Radiorchestra: Proactive management of millimeter-wave self-backhauled small cells via joint optimization of beamforming, user association, rate selection, and admission control,”IEEE Trans. Wireless Commun., vol. 22, no. 1, pp. 153–173, 2023

    2023

  52. [52]

    Radio resource management de- sign for RSMA: Optimization of beamforming, user admission, and discrete/continuous rates with imperfect SIC,

    L. F. Abanto-Leon, A. Krishnamoorthy, A. Garcia-Saavedra, G. H. Sim, R. Schober, and M. Hollick, “Radio resource management de- sign for RSMA: Optimization of beamforming, user admission, and discrete/continuous rates with imperfect SIC,”IEEE Trans. Mob. Comput., vol. 23, no. 12, pp. 11 498–11 518, 2024

    2024

  53. [53]

    Reversible instance normalization for accurate time-series forecasting against distribution shift,

    T. Kim, J. Kim, Y . Tae, C. Park, J.-H. Choi, and J. Choo, “Reversible instance normalization for accurate time-series forecasting against distribution shift,” inProc. 10th Int. Conf. Learn. Represent. (ICLR), 2022

    2022

  54. [54]

    Conformalized quantile regression,

    Y . Romano, E. Patterson, and E. Candes, “Conformalized quantile regression,” inProc. 33rd Adv. Neural Inf. Process. Syst. (NeurIPS), Jul 2019

    2019

  55. [55]

    A time series is worth 64 words: Long-term forecasting with transformers,

    Y . Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam, “A time series is worth 64 words: Long-term forecasting with transformers,” inProc. 11th Int. Conf. Learn. Represent. (ICLR), Apr. 2023, pp. 1–23

    2023

  56. [56]

    A MAPPO based scheme for joint resource allocation in UA V assisted MEC networks,

    M. Cheng, C. Zhu, M. Lin, and W.-P. Zhu, “A MAPPO based scheme for joint resource allocation in UA V assisted MEC networks,” inProc. IEEE/CIC Int. Conf. Commun. China (ICCC), Aug. 2024, pp. 42–47

    2024

  57. [57]

    Generative AI-driven reinforcement learning for beamforming and scheduling in multi-cell MIMO-NOMA systems,

    A. B. Adam, E. M. Diallo, M. S. A. Muthanna, R. I. Alkanhel, A. Muthanna, and M. Hammoudeh, “Generative AI-driven reinforcement learning for beamforming and scheduling in multi-cell MIMO-NOMA systems,”Phys. Commun., vol. 72, p. 102771, 2025

    2025

  58. [58]

    A denoising diffusion probabilistic model-based digital twinning of ISAC MIMO channel,

    J. Zhang, S. Xu, Z. Zhang, C. Li, and L. Yang, “A denoising diffusion probabilistic model-based digital twinning of ISAC MIMO channel,” IEEE Internet Things J., vol. 12, no. 15, pp. 29 121–29 134, 2025

    2025

  59. [59]

    Generative AI-enabled wireless communications for robust low-altitude economy networking,

    C. Zhao, J. Wang, R. Zhang, D. Niyato, G. Sun, H. Du, D. I. Kim, and A. Jamalipour, “Generative AI-enabled wireless communications for robust low-altitude economy networking,”IEEE Wireless Commun., pp. 1–9, 2025

    2025