Pith. sign in

REVIEW 2 major objections 4 minor 50 references

A large multimodal model builds a channel-capacity map from RGB-D and maps so handovers can be decided proactively, raising average capacity about 45% over 5G NR in dense mmWave simulations.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-14 15:38 UTC pith:NAVP3MTE

load-bearing objection Solid systems paper that turns an LMM into a joint trajectory/CCM/blockage engine for proactive UDN handover; gains are large in Sionna but rest on same-sim fine-tuning. the 2 major comments →

arxiv 2607.09795 v1 pith:NAVP3MTE submitted 2026-07-09 cs.IT cs.AIeess.SPmath.IT

Large Multimodal Model-Based Environment-Aware Mobility Management

Seokhyun Jeong , Sangmok Shin , Seungnyun Kim , Jiao Wu , Byonghyo Shim This is my paper
classification cs.IT cs.AIeess.SPmath.IT
keywords large multimodal modelmobility managementhandoverchannel capacity mapultra-dense networkmmWaveenvironment-awareproactive handover
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper argues that mobility management in ultra-dense mmWave networks fails when it reacts only to radio measurements, because channels jump discontinuously with reflections and blockages. The authors introduce LMM-EMM: a large multimodal model reads bird's-eye maps and SBS-view RGB-D images, extracts UE motion patterns and reflection geometry, and learns a channel capacity map (CCM) that maps UE/SBS positions plus environment to ideal and NLoS capacities. Future capacities along predicted trajectories are then refined for dynamic blockages and fed into a dynamic-programming solver that chooses the SBS sequence maximizing cumulative effective capacity (accounting for handover interruption). In ray-traced urban and suburban simulations the scheme delivers roughly 45% higher average capacity than standard 5G NR handover and 15-21% higher than LSTM- and DRL-based proactive baselines, while remaining inside beam-coherence time. A sympathetic reader cares because the same environmental context that humans use to anticipate coverage is shown to be extractable by an LMM and sufficient to replace real-time measurements for reliable proactive association.

Core claim

An LMM that jointly reasons over BEV maps, SBS-view RGB-D images and historical positions can construct a channel capacity map accurate enough that future capacities predicted from it, after blockage refinement, let dynamic programming select proactive handovers whose cumulative capacity substantially exceeds both reactive 5G NR and prior RF-only or vision-aided proactive schemes.

What carries the argument

Channel capacity map (CCM): the learned end-to-end mapping from UE position, SBS position and reflector geometry (implicitly encoded in a BEV image) to the static ideal and NLoS channel capacities; once built, it supplies the future rates that the DP handover optimizer maximizes.

Load-bearing premise

That a surrogate CCM fine-tuned from simulated BEV images and ray-tracing data remains accurate enough in real deployments that predicted capacities can safely replace live measurements for handover decisions.

What would settle it

Deploy the identical LMM-EMM pipeline on a live outdoor UDN testbed with measured RGB-D, GNSS trajectories and real mmWave CSI; if the measured average capacity gain over 5G NR handover falls well below the simulated 45% (or if CCM NMSE collapses under material or small-object mismatch), the central claim is falsified.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper proposes LMM-EMM for proactive mobility management in mmWave UDNs. An LMM (LLaVA-1.5-7B with LoRA) processes BEV maps and SBS-view RGB-D images to predict UE trajectories, infer multipath geometry, and construct a surrogate channel capacity map (CCM) ˜f_ccm (Eq. 26) that maps positions to ideal/NLoS capacities. Dynamic blockages are predicted from object trajectories; future capacities feed a DP solver (Eqs. 33–37) that maximizes cumulative effective capacity while penalizing handovers via µ. Simulations in Sionna RT claim ~45% average capacity gain over 5G NR and 15–21% over LSTM/DRL baselines at SNR=15 dB (Figs. 16–20), with supporting ablations on noise, material mismatch, soft blockage, and lightweight backbones.

Significance. If the gains transfer beyond the training simulator, the work would be a meaningful advance for environment-aware 6G mobility: it cleanly derives the piecewise-continuous CCM from geometric multipath parameters (Lemma 1, Remarks 1–3, Eqs. 4–18), formulates a proper cumulative-capacity DP problem that accounts for handover cost, and supplies unusually thorough sensitivity tables (I–III) plus multi-layout results (urban/suburban/highway/indoor). The latency analysis (Sec. IV-A) showing end-to-end delay under beam coherence time and the practical-issue discussion (noisy sensing, FoV, soft blockage) are concrete strengths that go beyond typical DL-handover papers.

major comments (2)
  1. [Sec. V-B, Eq. (26), Table II] Sec. V-B and Eq. (26): the surrogate CCM is obtained by supervised fine-tuning on 16 k/2 k/2 k samples generated by the identical Sionna RT + 3GPP TR 38.901 pipeline used for every evaluation scenario (Figs. 13, 16–22 and Tables I–III). This creates train–test distribution circularity. Table II already shows that additive CCM noise of variance 3 bps/Hz drops average capacity from 4.770 to 4.554 bps/Hz—the largest degradation among modules—yet all material-mismatch (Fig. 10), soft-blockage (Fig. 12) and small-object (Table III) ablations remain inside the same generative model. The central 45 %/21 %/15 % claims therefore rest on an untested assumption that ˜f_ccm recovers real multipath geometry (Eqs. 4–7). Either a cross-simulator or measurement-based validation, or a substantially expanded sim-to-real discussion quantifying the expected gap, is required before the gains can be regarded
  2. [Remark 3, Eq. (18), Sec. III-B] Remark 3 / Eq. (18) and Sec. III-B: the claim that a BEV image alone is a faithful surrogate for the full reflector set E (including Fresnel coefficients, roughness and permittivity that enter the LSFCs) is load-bearing for replacing real-time measurements by predicted capacities. The paper itself reports a 5.2 % capacity drop under material mismatch (Fig. 10) and only cosine similarity >0.93 under soft blockage (Fig. 12). Because the DP solution (Eqs. 35–37) is driven by these predicted capacities, an explicit bound or Monte-Carlo quantification of how CCM approximation error propagates into the optimality gap of the chosen SBS sequence is missing and necessary to support the “substantial” improvement language in the abstract and Sec. V.
minor comments (4)
  1. [Fig. 1] Fig. 1 caption contains the typo “Handover dicision”; correct to “decision”.
  2. [Sec. III-B] The instruction-prompt examples in Fig. 6 and Sec. III are helpful, but it is never stated how continuous capacity values are extracted from the LMM’s free-text response (regression head, constrained decoding, or post-processing). A short clarification would aid reproducibility.
  3. [Eq. (20)] Notation for time windows (Tw, Tp) and the handover indicator 1_ho is introduced cleanly, yet the effective-capacity expression (Eq. 20) re-uses m^(t-1:t) without restating the domain; a one-line reminder would improve readability.
  4. [Sec. I, Sec. II-D] Related-work discussion of channel-knowledge maps (Zeng et al.) and recent vision-aided handover papers is present but could more explicitly contrast the LMM’s shared environmental embedding against task-specific CNN/LSTM pipelines.

Circularity Check

1 steps flagged

Mild train-test circularity from fine-tuning LMM CCM surrogate on the identical Sionna RT ray-tracer used for all evaluation; no definitional or self-citation forcing of the central capacity gains.

specific steps
  1. fitted input called prediction [Sec. III-B Eq. (26) + Sec. V-B (dataset generation and SFT) + Table II]
    "we learn a surrogate function ˜f_ccm that takes the BEV map image, which implicitly encodes E: (R_ideal(m), R_nlos(m)) = ˜f_ccm(p_ue, p_sbs,m, I_bev). ... For LMM fine-tuning, we employ supervised fine-tuning (SFT) in which the LMM is trained on input–output pairs generated by a real-world wireless simulator (NVIDIA Sionna RT) ... The generated dataset for fine-tuning contains N_train=16000, N_val=2000, and N_test=2000 samples ... When the noise with a variance of 3 bps/Hz is added to the CCM estimation, the cumulative capacity per unit bandwidth decreases to 4.554 bps/Hz"

    The surrogate CCM is fitted by LoRA SFT exclusively to (position, capacity) pairs produced by the identical Sionna RT + 3GPP ray-tracer that later generates all test trajectories, urban/suburban layouts, material-mismatch ablations and capacity numbers. Consequently the reported NMSE and the end-to-end capacity gains that rely on those predicted capacities are statistically forced inside the training generative model; they do not constitute an independent prediction of real multipath geometry outside that simulator. (Held-out split and external baselines keep the circularity mild rather than definitional.)

full rationale

The paper's derivation chain is self-contained and non-circular in the strong sense: Lemma 1 and Remarks 1-3 derive that ideal/NLoS capacities are functions of LSFCs (hence of geometry E) under large-N and geometric multipath assumptions, with standard proofs; the surrogate ˜f_ccm of Eq. (26) is then an explicit LMM approximation of that mapping, trained by LoRA SFT on held-out Sionna pairs and used for trajectory-conditioned prediction + DP on the external cumulative-R_eff objective (23). End-to-end gains (Figs. 16-20, ~45% vs 5G NR, 15-21% vs LSTM/DRL) are measured against independent baselines that do not share LMM weights or the CCM fit. The only mild circularity is that all 16k/2k/2k samples, urban/suburban layouts, material-mismatch and small-object ablations, and few-shot adaptations remain inside the same generative model (Sionna RT + 3GPP TR 38.901) used for fine-tuning; thus reported CCM NMSE and capacity numbers cannot falsify the surrogate outside that simulator. This is ordinary ML-sim practice, not a definitional reduction or load-bearing self-citation uniqueness claim, so score remains low (2). No self-definitional equations, no uniqueness theorems imported from the authors' prior work to force the result, and no renaming of known patterns.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 1 invented entities

The central claim rests on standard geometric multipath models, the large-N capacity approximation, the assumption that BEV images plus LMM reasoning can surrogate the true reflector set E, and a handful of free parameters (LoRA rank, prediction horizon, handover cost) chosen for the experiments. No new physical entities are postulated; the CCM is an engineering construct rather than a fundamental object.

free parameters (4)
  • LoRA rank r = 16
    Set to 16 for fine-tuning LLaVA-1.5-7B; directly controls how much of the pretrained model is adapted to the wireless task.
  • prediction horizon Tp and observation window Tw = 5
    Both fixed to 5 time slots; determine how far ahead the LMM must forecast trajectories and capacities.
  • handover cost coefficient mu = tau_ho=36 ms (default)
    Defined from tau_ho / (tau_s - tau_ce); numerical value of interruption time (36 ms default) is chosen by the authors and strongly affects the DP objective.
  • learning rate schedule = 1e-4
    1e-4 decayed by 0.1 every 10 steps; controls convergence of the supervised fine-tuning loss.
axioms (4)
  • standard math When N is large, MISO capacity reduces to a function of large-scale fading coefficients only (Lemma 1 / law of large numbers).
    Standard asymptotic argument used to justify the CCM definition; invoked in Appendix A.
  • domain assumption Static channel capacities are fully determined by UE position, SBS position and reflector geometry E (Remark 3).
    Core modeling premise that allows the surrogate CCM to be learned from BEV images alone.
  • ad hoc to paper BEV map images plus LMM reasoning can serve as a faithful surrogate for the true reflector set E (Eq. 26).
    The paper never proves that the LMM recovers E; it only shows that the resulting capacity estimates are accurate enough on the simulated data.
  • domain assumption Hard LoS/NLoS indicator plus specular reflection model is adequate for mmWave capacity comparison (soft-blockage discussion in Sec. IV).
    Justified by short wavelength and high penetration loss; cosine-similarity checks are provided but remain simulation-based.
invented entities (1)
  • channel capacity map (CCM) no independent evidence
    purpose: End-to-end mapping from (UE position, SBS position, environment) to ideal and NLoS capacities that enables measurement-free future-capacity prediction.
    Engineering construct introduced by the authors; no independent physical existence claimed beyond the geometric model already assumed.

pith-pipeline@v1.1.0-grok45 · 31344 in / 3217 out tokens · 25090 ms · 2026-07-14T15:38:03.230255+00:00 · methodology

0 comments
read the original abstract

Recently, large language models (LLMs) have been successfully adopted in various fields, including wireless communications, robotics, and autonomous vehicles, owing to their outstanding adaptability and reasoning abilities. Despite their huge potential, the application of LLMs for mobility management is relatively scarce since it requires not only analyzing wireless measurements but also predicting dynamic user trajectories and making real-time handover decisions across densely deployed small base stations (SBSs). In this paper, we propose an environment-aware mobility management scheme based on large multimodal models (LMMs), which extend capabilities of LLMs to process multimodal sensing data. By leveraging LMMs, the proposed scheme extracts contextual information on the surrounding environments from RGB-D images to capture user equipment (UE) mobility patterns and identify signal reflections and blockages caused by static reflectors and dynamic obstacles. Using the extracted environmental information, the proposed scheme learns the intrinsic mapping from UE and SBS positions to channel capacity, referred to as channel capacity map (CCM), from which future channel capacities along UE trajectories are predicted. Based on the predicted channel capacities, we determine proactive handover decisions maximizing the cumulative channel capacities. Simulation results demonstrate that the proposed scheme achieves substantial channel capacity improvements over conventional deep learning (DL)-based approaches.

Figures

Figures reproduced from arXiv: 2607.09795 by Byonghyo Shim, Jiao Wu, Sangmok Shin, Seokhyun Jeong, Seungnyun Kim.

Figure 1
Figure 1. Figure 1: Visualization of mobility management in UDN systems. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Illustration of service interruption during the handover. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Overall procedure of the proposed LMM-EMM. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Illustration of LMM-based trajectory prediction using BEV [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Examples of instruction prompt and LMM response. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Illustration of LMM-based dynamic blockage prediction with [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: End-to-end latency vs. beam coherence time. • Static channel capacity estimation: T2 = 195 ms (195 ms for LMM inference) • Blockage prediction: T3 = 220 ms (25 ms for multi￾modal image capture and processing, and 195 ms for LMM inference) • Channel capacity refinement, report, and handover: T4 = 55 ms (25 ms for channel capacity refinement and report, and 30 ms for DP-based handover decision) Since blockag… view at source ↗
Figure 9
Figure 9. Figure 9: Examples of sensing images used for blockage prediction under noisy and low-resolution scenarios. Regarding the image resolution issue, objects that contribute to LoS blockage (e.g., car, bus, and truck) can still be effectively identified even in low-resolution images because they are typically large enough to occupy multiple pixels in the captured images. Furthermore, in our work, we crop and resize the … view at source ↗
Figure 10
Figure 10. Figure 10: Average channel capacity of LMM￾EMM with reflector material mismatch [PITH_FULL_IMAGE:figures/full_fig_p011_10.png] view at source ↗
Figure 13
Figure 13. Figure 13: Visualizations of the simulation environment on real-world scenarios. an autoregressive manner and performs handover based on the predicted capacity. 5) 5G NR handover [10]: A reactive handover scheme where the SBS determines the handover decisions based on consecutive RSRP reports from the UE. B. Multimodal Dataset Generation and LMM Fine-Tuning 1) Multimodal Dataset Generation: To construct a realistic … view at source ↗
Figure 14
Figure 14. Figure 14: Comparison of CCMs generated by LMM-EMM and the conventional DL-based scheme. [PITH_FULL_IMAGE:figures/full_fig_p012_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Visual comparison of predicted UE trajectories for [PITH_FULL_IMAGE:figures/full_fig_p012_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Average channel capacity as a func￾tion of SNR [PITH_FULL_IMAGE:figures/full_fig_p013_16.png] view at source ↗
Figure 19
Figure 19. Figure 19: Average channel capacity as a func￾tion of handover interruption time [PITH_FULL_IMAGE:figures/full_fig_p014_19.png] view at source ↗
Figure 22
Figure 22. Figure 22: Average channel capacity as a func￾tion of SNR in different architectural layouts after few-shot adaptation. 0 5 10 15 20 25 30 SNR [dB] 1 2 3 4 5 6 7 Average channel capacity [bps/Hz] Proposed LMM-EMM LMM-EMM w/ CNN+LSTM UE trajectory prediction LMM-EMM w/ CNN+LSTM channel capacity estimation LMM-EMM w/ CNN+LSTM blockage prediction [PITH_FULL_IMAGE:figures/full_fig_p015_22.png] view at source ↗

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

50 extracted references · 2 linked inside Pith

  1. [1]

    Language models are few-shot learners,

    T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askellet al., “Language models are few-shot learners,” inProc. Adv. Neural Inf. Process. Syst. (NeurIPS), vol. 33, 2020, pp. 1877–1901

    2020

  2. [2]

    Visual instruction tuning,

    H. Liu, C. Li, Q. Wu, and Y . J. Lee, “Visual instruction tuning,” inProc. Adv. Neural Inf. Process. Syst. (NeurIPS), vol. 36, 2023, pp. 34 892– 34 916

    2023

  3. [3]

    Chain-of-thought prompting elicits reasoning in large language models,

    J. Wei, X. Wang, D. Schuurmans, M. Bosma, F. Xia, E. Chi, Q. V . Le, D. Zhouet al., “Chain-of-thought prompting elicits reasoning in large language models,” inProc. Adv. Neural Inf. Process. Syst. (NeurIPS), vol. 35, 2022, pp. 24 824–24 837

    2022

  4. [4]

    Large language models for telecom: Forthcoming impact on the indus- try,

    A. Maatouk, N. Piovesan, F. Ayed, A. De Domenico, and M. Debbah, “Large language models for telecom: Forthcoming impact on the indus- try,”IEEE Commun. Mag., vol. 63, no. 1, pp. 62–68, 2025

    2025

  5. [5]

    Large multimodal model-empowered task-oriented autonomous communications: Design methodology and implementation challenges,

    H. J. Yang, H. Kim, H. Noh, S. Kim, and B. Shim, “Large multimodal model-empowered task-oriented autonomous communications: Design methodology and implementation challenges,”IEEE Veh. Technol. Mag., 2026. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. XX, NO. YY , ZZZZZZ 2026 16

    2026

  6. [6]

    Sensing and computer vision-aided mobility management for 6G millimeter and terahertz communication systems,

    Y . Ahn, J. Kim, S. Kim, S. Kim, and B. Shim, “Sensing and computer vision-aided mobility management for 6G millimeter and terahertz communication systems,”IEEE Trans. Commun., vol. 72, no. 10, pp. 6044–6058, 2024

    2024

  7. [7]

    Energy-efficient user association in mmWave/THz ultra-dense network via multi-agent deep reinforce- ment learning,

    J. Moon, S. Kim, H. Ju, and B. Shim, “Energy-efficient user association in mmWave/THz ultra-dense network via multi-agent deep reinforce- ment learning,”IEEE Trans. Green Commun. Netw., vol. 7, no. 2, pp. 692–706, 2023

    2023

  8. [8]

    Efficient channel probing and phase shift control for mmWave reconfigurable intelligent surface-aided communi- cations,

    S. Kim, J. Wu, and B. Shim, “Efficient channel probing and phase shift control for mmWave reconfigurable intelligent surface-aided communi- cations,”IEEE Trans. Wireless Commun., vol. 23, no. 1, pp. 231–246, 2023

    2023

  9. [9]

    Ultra-dense networks: A survey,

    M. Kamel, W. Hamouda, and A. Youssef, “Ultra-dense networks: A survey,”IEEE Commun. Surveys Tuts., vol. 18, no. 4, pp. 2522–2545, 2016. [10]Radio Resource Control (RRC), 3rd Generation Partnership Project 3GPP™ TS 38.331 V18.4.0, Dec. 2024, Release 18

    2016

  10. [10]

    A statistical theory of mobile-radio reception,

    R. H. Clarke, “A statistical theory of mobile-radio reception,”Bell Sys. Technol. J., vol. 47, no. 6, pp. 957–1000, 1968

    1968

  11. [11]

    Machine learning-based handovers for sub-6 GHz and mmWave integrated vehicular networks,

    L. Yan, H. Ding, L. Zhang, J. Liu, X. Fang, Y . Fang, M. Xiao, and X. Huang, “Machine learning-based handovers for sub-6 GHz and mmWave integrated vehicular networks,”IEEE Trans. Wireless Commun., vol. 18, no. 10, pp. 4873–4885, 2019

    2019

  12. [12]

    Multi-cell multi-beam prediction us- ing auto-encoder LSTM for mmWave systems,

    S. H. A. Shah and S. Rangan, “Multi-cell multi-beam prediction us- ing auto-encoder LSTM for mmWave systems,”IEEE Trans. Wireless Commun., vol. 21, no. 12, pp. 10 366–10 380, 2022

    2022

  13. [13]

    Mobility management paradigm shift: from reactive to proactive handover using AI/ML,

    H.-S. Park, H. Kim, C. Lee, and H. Lee, “Mobility management paradigm shift: from reactive to proactive handover using AI/ML,”IEEE Netw., vol. 38, no. 2, pp. 18–25, 2024

    2024

  14. [14]

    Enabling efficient blockage-aware handover in RIS-assisted mmWave cellular networks,

    L. Jiao, P. Wang, A. Alipour-Fanid, H. Zeng, and K. Zeng, “Enabling efficient blockage-aware handover in RIS-assisted mmWave cellular networks,”IEEE Trans. Wireless Commun., vol. 21, no. 4, pp. 2243– 2257, 2021

    2021

  15. [15]

    Self-adapting handover parameters optimization for SDN- enabled UDN,

    W. Huang, M. Wu, Z. Yang, K. Sun, H. Zhang, and A. Nal- lanathan, “Self-adapting handover parameters optimization for SDN- enabled UDN,”IEEE Trans. Wireless Commun., vol. 21, no. 8, pp. 6434–6447, 2022

    2022

  16. [16]

    Proactive handover type prediction and parameter optimization based on machine learning,

    K. Sun, Q. Han, Z. Yang, W. Huang, H. Zhang, and V . C. Leung, “Proactive handover type prediction and parameter optimization based on machine learning,”IEEE Trans. Wireless Commun., vol. 24, no. 4, pp. 3515–3528, 2025

    2025

  17. [17]

    Role of sensing and computer vision in 6G wireless communications,

    S. Kim, J. Moon, J. Kim, Y . Ahn, D. Kim, S. Kim, K. Shim, and B. Shim, “Role of sensing and computer vision in 6G wireless communications,” IEEE Wireless Commun., vol. 31, no. 5, pp. 264–271, 2024

    2024

  18. [18]

    Vision-aided 6G wireless communications: Blockage prediction and proactive handoff,

    G. Charan, M. Alrabeiah, and A. Alkhateeb, “Vision-aided 6G wireless communications: Blockage prediction and proactive handoff,”IEEE Trans. Veh. Technol., vol. 70, no. 10, pp. 10 193–10 208, 2021

    2021

  19. [19]

    Radar aided proactive blockage pre- diction in real-world millimeter wave systems,

    U. Demirhan and A. Alkhateeb, “Radar aided proactive blockage pre- diction in real-world millimeter wave systems,” inProc. IEEE Int. Conf. Commun. (ICC), 2022, pp. 4547–4552

    2022

  20. [20]

    Vision-aided blockage prediction and proactive handover for indoor mmWave and terahertz communica- tions,

    Y . Liu, J. Wu, S. Kim, and B. Shim, “Vision-aided blockage prediction and proactive handover for indoor mmWave and terahertz communica- tions,” inProc. IEEE Global Commun. Conf. (GLOBECOM), 2023, pp. 7411–7416

    2023

  21. [21]

    Joint sensing, communication, and AI: A trifecta for resilient THz user experiences,

    C. Chaccour, W. Saad, M. Debbah, and H. V . Poor, “Joint sensing, communication, and AI: A trifecta for resilient THz user experiences,” IEEE Trans. Wireless Commun., vol. 23, no. 9, pp. 11 444–11 460, 2024

    2024

  22. [22]

    Networked ISAC based UA V tracking and handover towards low-altitude economy,

    Y . Feng, C. Zhao, H. Luo, F. Gao, F. Liu, and S. Jin, “Networked ISAC based UA V tracking and handover towards low-altitude economy,”IEEE Trans. Wireless Commun., vol. 24, no. 9, pp. 7670–7685, 2025

    2025

  23. [23]

    A non-stochastic learning approach to energy efficient mobility management,

    C. Shen, C. Tekin, and M. van der Schaar, “A non-stochastic learning approach to energy efficient mobility management,”IEEE J. Sel. Areas Commun., vol. 34, no. 12, pp. 3854–3868, 2016

    2016

  24. [24]

    Capacity of MIMO systems based on measured wireless channels,

    A. F. Molisch, M. Steinbauer, M. Toeltsch, E. Bonek, and R. S. Thoma, “Capacity of MIMO systems based on measured wireless channels,” IEEE J. Sel. Areas Commun., vol. 20, no. 3, pp. 561–569, 2002

    2002

  25. [25]

    Utilizing large language models for advanced optimization and intelligent management in space-air-ground integrated networks,

    J. Tang, F. Tang, S. Long, M. Zhao, and N. Kato, “Utilizing large language models for advanced optimization and intelligent management in space-air-ground integrated networks,”IEEE Netw., vol. 39, no. 5, pp. 173–181, 2024

    2024

  26. [26]

    Characterizing the impact of dif- fuse scattering in urban millimeter-wave deployments,

    D. Solomitckii, Q. C. Li, T. Balercia, C. R. Da Silva, S. Talwar, S. Andreev, and Y . Koucheryavy, “Characterizing the impact of dif- fuse scattering in urban millimeter-wave deployments,”IEEE Wireless Commun. Lett., vol. 5, no. 4, pp. 432–435, 2016

    2016

  27. [27]

    Large multimodal model-based environment-aware channel estimation,

    S. Kim, S. Jeong, J. Wu, B. Shim, and M. Z. Win, “Large multimodal model-based environment-aware channel estimation,”IEEE J. Sel. Areas Commun., vol. 43, no. 12, pp. 4059–4075, 2025

    2025

  28. [28]

    Transformer-assisted parametric CSI feedback for mmWave massive MIMO systems,

    H. Ju, S. Jeong, S. Kim, B. Lee, and B. Shim, “Transformer-assisted parametric CSI feedback for mmWave massive MIMO systems,”IEEE Trans. Wireless Commun., vol. 23, no. 12, pp. 18 774–18 787, 2024. [30]Evolved Universal Terrestrial Radio Access Network (E-UTRAN); X2 Application Protocol (X2AP), 3rd Generation Partnership Project 3GPP™ TS 36.423 V18.4.0, Ma...

    2024

  29. [29]

    Channel prediction using adaptive bidirectional GRU for underwater MIMO communica- tions,

    X. Hu, Y . Huo, X. Dong, F.-Y . Wu, and A. Huang, “Channel prediction using adaptive bidirectional GRU for underwater MIMO communica- tions,”IEEE Internet Things J., vol. 11, no. 2, pp. 3250–3263, 2023

    2023

  30. [30]

    Proactive mobility management with trajectory prediction based on virtual cells in ultra-dense networks,

    Q. Liu, G. Chuai, J. Wang, and J. Pan, “Proactive mobility management with trajectory prediction based on virtual cells in ultra-dense networks,” IEEE Trans. Veh. Technol., vol. 69, no. 8, pp. 8832–8842, 2020

    2020

  31. [31]

    A tutorial on 5G positioning,

    L. Italiano, B. C. Tedeschini, M. Brambilla, H. Huang, M. Nicoli, and H. Wymeersch, “A tutorial on 5G positioning,”IEEE Commun. Surveys Tuts., vol. 27, no. 3, pp. 1488–1535, 2024

    2024

  32. [32]

    Vision-aided positioning and beam focusing for 6G terahertz communications,

    S. Kim, J. Moon, J. Wu, B. Shim, and M. Z. Win, “Vision-aided positioning and beam focusing for 6G terahertz communications,”IEEE J. Sel. Areas Commun., vol. 42, no. 9, pp. 2503–2519, 2024

    2024

  33. [33]

    A tutorial on environment-aware communications via channel knowledge map for 6G,

    Y . Zeng, J. Chen, J. Xu, D. Wu, X. Xu, S. Jin, X. Gao, D. Gesbert, S. Cui, and R. Zhang, “A tutorial on environment-aware communications via channel knowledge map for 6G,”IEEE Commun. Surveys Tuts., vol. 26, no. 3, pp. 1478–1519, 2024

    2024

  34. [34]

    Traj-LLM: A new exploration for empowering trajectory prediction with pre-trained large language models,

    Z. Lan, L. Liu, B. Fan, Y . Lv, Y . Ren, and Z. Cui, “Traj-LLM: A new exploration for empowering trajectory prediction with pre-trained large language models,”IEEE Trans. Intell. Veh., vol. 10, no. 2, pp. 794–807, 2025

    2025

  35. [35]

    Deep learning- based NLoS localization using geometric map image,

    I. Keum, J. Son, H. Kim, J. Moon, and B. Shim, “Deep learning- based NLoS localization using geometric map image,”IEEE Trans. Veh. Technol., 2025

    2025

  36. [36]

    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. Int. Conf. Learn. Representations (ICLR), 2023

    2023

  37. [37]

    Large multimodal model-based environment-aware beam management,

    S. Kim, S. Saha, S. Jeong, B. Shim, and M. Z. Win, “Large multimodal model-based environment-aware beam management,”IEEE J. Sel. Areas Commun., vol. 44, pp. 991–1007, 2026

    2026

  38. [38]

    Dynamic programming,

    R. Bellman, “Dynamic programming,”Science, vol. 153, no. 3731, pp. 34–37, 1966

    1966

  39. [39]

    The impact of beamwidth on temporal channel variation in vehicular channels and its implications,

    V . Va, J. Choi, and R. W. Heath, “The impact of beamwidth on temporal channel variation in vehicular channels and its implications,”IEEE Trans. Veh. Technol., vol. 66, no. 6, pp. 5014–5029, 2016

    2016

  40. [40]

    Multi- weather image restoration via domain translation,

    P. W. Patil, S. Gupta, S. Rana, S. Venkatesh, and S. Murala, “Multi- weather image restoration via domain translation,” inProc. Int. Conf. Comput. Vis. (ICCV), 2023, pp. 21 696–21 705

    2023

  41. [41]

    Openstreetmap: User-generated street maps,

    M. Haklay and P. Weber, “Openstreetmap: User-generated street maps,” IEEE Pervasive Comput., vol. 7, no. 4, pp. 12–18, 2008. [44]SMPTE Engineering Guideline - SD-SDI and HD-SDI Standards Roadmap, IEEE Standard EG 2111-1:2021, 2021

    2008

  42. [42]

    Millimeter- wave human blockage at 73 ghz with a simple double knife-edge diffraction model and extension for directional antennas,

    G. R. MacCartney, S. Deng, S. Sun, and T. S. Rappaport, “Millimeter- wave human blockage at 73 ghz with a simple double knife-edge diffraction model and extension for directional antennas,” inIEEE Veh. Technol. Conf. (VTC-Fall), 2016

    2016

  43. [43]

    Sionna RT: Differentiable ray tracing for radio propagation modeling,

    J. Hoydis, F. A ¨ıt Aoudia, S. Cammerer, M. Nimier-David, N. Binder, G. Marcus, and A. Keller, “Sionna RT: Differentiable ray tracing for radio propagation modeling,” inProc. IEEE Global Commun. Conf. Workshops (GC Wkshps), 2023, pp. 317–321. [47]Study on channel model for frequencies from 0.5 to 100 GHz, 3rd Generation Partnership Project 3GPP™ TR 38.901...

    2023

  44. [44]

    LoRA: Low-rank adaptation of large language models,

    E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, W. Chenet al., “LoRA: Low-rank adaptation of large language models,” inProc. Int. Conf. Learn. Representations (ICLR), 2022

    2022

  45. [45]

    Toward environment-aware 6G communications via channel knowledge map,

    Y . Zeng and X. Xu, “Toward environment-aware 6G communications via channel knowledge map,”IEEE Wireless Commun., vol. 28, no. 3, pp. 84–91, 2021

    2021

  46. [46]

    Mobilevlm: A fast, strong and open vision language assistant for mobile devices,

    X. Chuet al., “Mobilevlm: A fast, strong and open vision language assistant for mobile devices,”arXiv preprint arXiv:2312.16886, 2023

    Pith/arXiv arXiv 2023

  47. [47]

    Tinyllava: A framework of small-scale large multimodal models,

    B. Zhouet al., “Tinyllava: A framework of small-scale large multimodal models,”arXiv preprint arXiv:2402.14289, 2024

    Pith/arXiv arXiv 2024

  48. [48]

    Proactive received power prediction using machine learning and depth images for mmWave networks,

    T. Nishioet al., “Proactive received power prediction using machine learning and depth images for mmWave networks,”IEEE J. Sel. Areas Commun., vol. 37, no. 11, pp. 2413–2427, 2019

    2019

  49. [49]

    Deep learning for mmWave beam and blockage prediction using sub-6 GHz channels,

    M. Alrabeiah and A. Alkhateeb, “Deep learning for mmWave beam and blockage prediction using sub-6 GHz channels,”IEEE Trans. Commun., vol. 68, no. 9, pp. 5504–5518, 2020

    2020

  50. [50]

    Efficient large multi-modal models via visual context compression,

    J. Chen, L. Ye, J. He, Z.-Y . Wang, D. Khashabi, and A. Yuille, “Efficient large multi-modal models via visual context compression,” inProc. Adv. Neural Inf. Process. Syst. (NeurIPS), vol. 37, 2024, pp. 73 986–74 007. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. XX, NO. YY , ZZZZZZ 2026 17 Seokhyun Jeong(Student Member, IEEE) received the B.S. degre...

    2024