Pith. sign in

REVIEW 3 major objections 5 minor 61 references

Towards Wireless Native Big AI Model: The Mission and Approach Differ From Large Language Model

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper argues that general wireless AI requires a wireless-native big model that learns electromagnetic laws directly, rather than adapting large language models.

desk verdict Useful roadmap with a novel human-centric vs hyper-cognitive framing, but the core inference that multi-task generalization implies internalizing EM laws is asserted rather than shown. read the letter →

arxiv 2412.09041 v3 pith:K27JSMEQ submitted 2024-12-12 eess.SP cs.SYeess.SY

Zirui Chen , Zhaoyang Zhang , Chenyu Liu , Ziqing Xing This is my paper · ORCID
classification eess.SPcs.SYeess.SY
keywords wirelessnativeAIbigmodelfoundation6Gelectromagneticlawshyper-cognitiveintelligencescalingphysics-drivenlearning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that the route to general wireless AI is not to adapt large language models, but to build a wireless-native big AI model (wBAIM) that learns from electromagnetic phenomena directly. It claims language intelligence and wireless intelligence are different in kind: language is human-centric, while electromagnetic systems are hyper-cognitive, meaning they exceed direct human understanding. Because wireless tasks and scenarios share underlying electromagnetic laws, a single model that generalizes across tasks must be compressing those laws. The paper surveys current paradigms, identifies the peculiarities of wireless data, models, and applications, and proposes four methodological pillars: hybrid data collection, a physics-driven learning paradigm, wireless scaling laws, and structural prompting. If the vision is right, 6G wireless intelligence would shift from task-specific models to a unified foundation model that also supports radar and remote sensing.

What carries the argument

The load-bearing distinction is between human-centric and hyper-cognitive intelligence, coupled with the identity that cross-task, cross-scenario generalization implies learning common electromagnetic laws. The paper's proposed machinery for realizing this is a synthesis of four LLM evolutionary drivers mapped to wireless: hybrid data collection that mixes multi-scenario measurements with simulation; a physics-driven learning paradigm that injects known electromagnetic structure into model design; wireless scaling laws in which jointly processing multimodal, multi-user, multi-scenario data is what produces the 'more is different' emergence effect; and structural prompting, where prompts are strictly structured multimodal data rather than free-form language, so the model can adapt without retraining.

What would settle it

Run a controlled scaling study on real measured wireless channel data: train a family of wBAIMs with increasing parameter counts and increasingly diverse multi-scenario training sets, and check whether performance on an unseen scenario improves smoothly or abruptly. If no such improvement appears, or if gains saturate, the paper's wireless scaling-law premise is falsified.

Watch

Extended reading notes

Core claim

The central claim is that a wireless-native big AI model should be developed as a hyper-cognitive model whose training signal is observed electromagnetic phenomena rather than human-generated text. On this view, the reason LLMs cannot be transplanted to wireless is not missing domain knowledge but a mismatch in intelligence orientation: language models imitate human cognition, whereas electromagnetic systems require intelligence that surpasses it. The paper asserts that if a single model generalizes across multiple wireless tasks and scenarios, it must inherently capture the commonalities among them, namely the general electromagnetic laws, which it calls 'compression is intelligence' in the wireless context. From this it follows that wBAIM's defining features of multi-task integration, multi-scenario unification, and all-in-one scheduling are not conveniences but implicit evidence that the model has learned real physics.

Load-bearing premise

The load-bearing premise is that big-model scaling and joint multi-modal, multi-user, multi-scenario training will generate the same 'more is different' emergence in wireless models that it did in language models; the paper presents this as a principle with only preliminary confirmation and no scaling-law fit.

Editorial extensions

If this is right

  • LLM-based wireless approaches, including wireless LLMs and LLM-based agents, will remain limited to interaction and simple tasks; hard wireless problems such as high-rate transmission and scenario sensing will need a native model.
  • Training wBAIM will require mixing measured and simulated data, with simulation providing high-fidelity multipath components and unbiased user locations that are difficult to obtain from real measurements.
  • Adaptation of wBAIM will rely on structural prompts rather than fine-tuning, because retraining is too slow and resource-hungry for wireless nodes with frequently changing environments.
  • A successful wBAIM should be usable beyond communications, providing technical support to remote sensing and radar systems.
  • Collaborating wBAIM with AI-agent modules for planning, observation, memory, and small tool libraries would turn it into an active intelligent wireless brain rather than a passive predictor.

Reading between the lines

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

  • If the compression-is-intelligence claim holds for electromagnetics, then wBAIM performance on a held-out task becomes a measurable proxy for whether the model has internalized physical laws; one could test this by probing latent representations against analytic channel models.
  • The paper's wireless scaling-law argument could be made testable by measuring whether multi-scenario pre-training produces error drops on new scenarios that fit a power law in model size and data mixing ratio, but the paper does not provide such exponents.
  • The hyper-cognitive framing suggests that wBAIM, if successful, would be a template for other AI-for-science domains such as weather, materials, or fluid dynamics, where the teacher is the system itself rather than human labels.
  • The decentralized constraint implies that wBAIM pre-training may need to proceed through model-interaction schemes like federated learning, which the paper mentions only briefly.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. This position paper argues that applying large language model (LLM) technology to wireless systems is insufficient and that a wireless-native big AI model (wBAIM) should be developed instead. It reviews 36 publications on BAIM for wireless, categorizes them into wireless LLM, LLM-based wireless agents, and wBAIM, and proposes that wireless intelligence is fundamentally different from language intelligence because it is hyper-cognitive rather than human-centric. The paper then identifies peculiarities of wBAIM in data attributes, model functionality, and applied scenarios, and proposes five methodologies: hybrid data collection, physics-driven learning, wireless scaling laws, structural prompting, and collaboration with AI agents. The central thesis is that wBAIM should learn electromagnetic laws directly from wireless data and will generalize across tasks and scenarios as a result.

Significance. If the central thesis holds, this paper would provide a useful conceptual foundation for a shift from task-specific wireless AI models to unified foundation models for 6G and electromagnetic sensing. The paper's strengths are its clear research questions, systematic taxonomy of current BAIM-for-wireless paradigms, and explicit identification of differences between language and wireless intelligence. It also gives specific methodological directions, such as structural prompts and physics-driven learning, that are actionable. The paper is honest about the early stage of wBAIM research and does not overclaim experimental results. Its main weakness is that the logical link between cross-task generalization and internalization of electromagnetic laws is asserted rather than demonstrated, and the scaling-law argument rests on analogy to LLMs without wireless-specific evidence.

major comments (3)
  1. [Section 3.2] The statement that 'If a single model can be used to generalize across multiple wireless tasks and scenarios, it must inherently capture the commonalities among them—that is, the general electromagnetic laws' is an unsupported logical inference. Generalization across tasks and scenarios can arise from shared but non-physical feature statistics, task-specific decoders over a domain-agnostic representation, or dataset biases such as scenario identifiers and common measurement artifacts. The paper provides no experimental or analytical evidence that the commonalities learned by wireless models are physical laws. Since this equivalence is used to justify the wBAIM definition and its implicit evaluation metrics, it is load-bearing and needs to be either supported with evidence (for example, probing experiments or out-of-distribution tests that distinguish physics-consistent predictions from statistical shortcuts) or reformulated as a hypothesis rather than a necessity.
  2. [Section 5.3] The claim that 'increasing model size is also essential for the performance of wireless models, especially in terms of generalization' and that jointly processing multi-modal, multi-user, multi-scenario data will generate the 'more is different' emergence effect is asserted without wireless-specific evidence. The paper cites reference [19] as preliminary confirmation, but no scaling law, fitted exponent, or experimental demonstration is provided. For a position paper it is acceptable to propose scaling as a hypothesis, but the current wording states it as a principle and therefore overstates the support. This is load-bearing because the entire methodology of building a large wBAIM depends on this transfer of scaling behavior from LLMs to wireless models.
  3. [Section 4.2] There is a tension between the claim in Section 4.2 that 'the fundamental laws of electromagnetic waves are relatively simple' and the claim in Section 3.2 that a single model generalizing across tasks and scenarios must capture these laws. If the laws are simple, it is not obvious why very large models and massive pre-training are necessary to learn them; compact physics-informed models might suffice. The paper should reconcile these statements by explaining why, despite the simplicity of the underlying laws, scale and data diversity are still needed—for example, because the complexity lies in the coupling with scattering environments and transmission mechanisms. As written, the two sections pull in opposite directions and weaken the scaling argument.
minor comments (5)
  1. [Figure 3] The statistical text under Figure 3 contains a clear typesetting artifact ('/uni00000033/uni00000055/...') that renders as garbled code instead of readable labels; this needs to be fixed in the production version.
  2. [Title page] The title page says 'POSITION P APER' with an extra space; also, the corresponding author email is written as 'ning ming@zju.edu.cn', which appears to be a spacing error.
  3. [Abstract and Introduction] The acronym 'BAIM' is used both as a general term for big AI models and as part of 'wBAIM'; the paper would benefit from a brief definition of BAIM in the abstract or first use, since the abstract introduces it without explanation.
  4. [Section 4.1] The phrase 'L VMs' contains an unintended space; it should read 'LVMs'.
  5. [References] Reference [34] contains the typo 'telecom-specfic'; please correct to 'telecom-specific'.

Circularity Check

2 steps flagged · score 4.0 of 10

The wBAIM significance argument equates multi-task generalization with capturing electromagnetic laws by definition, and the wireless scaling principle leans on the authors' own prior paper.

  1. self definitional [Section 3.2, Scientific Significance of wBAIM]
    "If a single model can be used to generalize across multiple wireless tasks and scenarios, it must inherently capture the commonalities among them—that is, the general electromagnetic laws. In other words, this is the 'compression is intelligence' in the wireless context."

    The paper first says implicit assessment metrics are essential because black-box models cannot be directly checked for EM-law competence, then cites [19] as defining wBAIM by multi-task integration, multi-scenario unification, and all-in-one scheduling. The quoted sentence then treats cross-task/cross-scenario generalization as entailing capture of 'commonalities,' and glosses those commonalities as 'the general electromagnetic laws.' This makes the conclusion that wBAIM internalizes electromagnetic laws true by stipulation: the chosen operational metric for law-capture is multi-task generalization.

  2. self citation load bearing [Section 5.3, Wireless Scaling Laws]
    "This principle is preliminarily confirmed in [19]."

    The preceding sentences argue that 'increasing model size is also essential for the performance of wireless models' and that sufficient neurons are needed for storing knowledge and reasoning. The only cited confirmation is reference [19], authored by two of the current paper's authors (Chen and Zhang). No independent scaling law, fitted exponent, or experimental result is provided in the present paper. This self-citation is load-bearing for the 'wireless scaling laws' methodology pillar, though it is subordinate to the paper's broader roadmap and does not by itself force the central wireless-native conclusion.

full rationale

This is a position paper rather than a derivational or empirical study, so most of its claims are arguments, comparisons, and methodological suggestions rather than predictions that could reduce to fitted inputs. The literature review and wireless-vs-language comparisons contain substantial independent content. The main circularity risk is in Section 3.2, where the paper's assertion that a model generalizing across wireless tasks and scenarios 'must inherently capture' electromagnetic laws is made true by how the implicit assessment metric is chosen: multi-task and multi-scenario generalization is effectively defined as evidence of EM-law capture, with 'commonalities' equated to 'general electromagnetic laws.' That is a self-definitional move, though it supports the paper's framing of wBAIM's significance rather than a quantitative result. A second, milder issue is the reliance on the authors' own prior work [19] to 'preliminarily confirm' the wireless scaling-law principle, with no external evidence supplied. These issues do not make the entire roadmap circular: the paper's proposed methodologies—hybrid data collection, physics-driven learning, structural prompting, agent collaboration—are independent of the definitional shortcut. Overall, some self-citation and a definitional leap are present, but the central claim retains independent content, warranting a moderate rather than severe circularity score.

Assumptions & free parameters 0 free parameters · 5 assumptions · 1 invented entities

The central argument rests on several unverified background assumptions, principally that LLM-like scaling behavior will emerge in wireless models and that a single model generalizing across wireless tasks must have learned electromagnetic laws. The wBAIM concept is adopted from the authors' own prior work, adding to circularity burden.

assumptions (5)
  • domain assumption LLM success is driven by four elements: extensive data, a general-intelligence-oriented learning paradigm, emergence from scale, and efficient adaptation.
    Section 5 states this as the basis for mapping language evolution to wireless, but it is an interpretation, not a proven law.
  • ad hoc to paper A model that generalizes across multiple wireless tasks and scenarios must have captured the common electromagnetic laws ('compression is intelligence').
    Section 3.2 introduces this equivalence without independent evidence; it is the paper's own epistemic claim.
  • domain assumption Increasing wireless model scale yields performance gains and an emergence effect analogous to LLMs.
    Section 5.3 asserts this principle and cites [19] as preliminary confirmation, but no scaling law is fitted or shown.
  • domain assumption Wireless data have strict logical order, correspondence, and structure, so prompts must be structural rather than linguistic.
    Sections 4.1 and 5.4 argue this from examples (CSI, modulation symbols, positioning), but it is a general claim about all wireless data.
  • ad hoc to paper The human-centric versus hyper-cognitive dichotomy is a valid basis for understanding AI domains.
    Section 3.1 introduces this taxonomy; it is a framing device, not an empirically validated characterization.
invented entities (1)
  • wBAIM (wireless-native big AI model)
    purpose: A proposed class of foundation model that integrates multi-task, multi-scenario, and all-in-one scheduling for wireless systems.
    Introduced in the authors' prior work [19] and adopted here as the central concept. Only a small demonstrative case study exists in [19]; no full implementation or falsifiable prediction is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Wireless Native Big AI Model: The Mission and Approach Differ From Large Language Model." pith.science (2026). https://pith.science/paper/K27JSMEQ

@misc{pith2026241209041,
  author       = {Pith},
  title        = {Pith review of: Towards Wireless Native Big AI Model: The Mission and Approach Differ From Large Language Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K27JSMEQ}},
  note         = {Machine review of arXiv:2412.09041}
}
read the original abstract

Research on leveraging big artificial intelligence model (BAIM) technology to drive the intelligent evolution of wireless networks is emerging. However, breakthroughs in generalization brought about by BAIM techniques mainly occur in natural language processing. There is a lack of a clear technical direction on how to efficiently apply BAIM techniques to wireless systems, which typically have many additional peculiarities. To this end, this paper reviews recent research on BAIM for wireless systems and assesses the current state of the field. It then analyzes and compares the differences between language intelligence and wireless intelligence on multiple levels, including scientific foundations, core usages, and technical details. It highlights the necessity and scientific significance of developing wireless native BAIM technologies, as well as specific issues that need to be considered for technical implementation. Finally, by synthesizing the evolutionary laws of language models with the particularities of wireless systems, this paper provides several instructive methodologies for developing wireless native BAIM.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

61 extracted references · 35 canonical work pages

  1. [19]

    Big AI models for 6G wireless networks: Opportunities, challenges, and research directions

    Chen Z R, Zhang Z Y, Yang Z H. Big AI models for 6G wireless networks: Opportunities, challenges, and research directions. IEEE Wireless Commun, 2024, 31: 164-172

    work page 2024

  2. [1]

    AI for 5G : research directions and paradigms

    You X H, Zhang C, Tan X S, et al. AI for 5G : research directions and paradigms. Sci China Inf Sci, 2019, 62: 21301

    work page 2019

  3. [2]

    Edge learning for B5G networks with distributed signal processing: Semantic communication, edge computing, and wireless sensing

    Xu W, Yang Z H, Ng D W K, et al. Edge learning for B5G networks with distributed signal processing: Semantic communication, edge computing, and wireless sensing. IEEE J Sel Top Sign Proces, 2023. 17: 9-39

    work page 2023

  4. [3]

    Energy efficient semantic communication over wireless networks with rate splitting

    Yang Z H, Chen M Z, Zhang Z Y, et al. Energy efficient semantic communication over wireless networks with rate splitting. IEEE J Sel Areas Commun, 2023, 41: 1484-1495

    work page 2023

  5. [4]

    The roadmap to 6G: AI empowered wireless networks

    Letaief K B, Chen W, Shi Y M, et al. The roadmap to 6G: AI empowered wireless networks. IEEE Commun Mag, 2019. 57: 84-90

    work page 2019

  6. [5]

    When AI meets sustainable 6G

    You X, Huang Y, Zhang C, et al. When AI meets sustainable 6G. Sci China Inf Sci, 2025, 68: 110301

    work page 2025

  7. [6]

    Pushing AI to wireless network edge: an overview on integrated sensing, communication, and computation towards 6G

    Zhu G, Lyu Z, Jiao X, et al. Pushing AI to wireless network edge: an overview on integrated sensing, communication, and computation towards 6G. Sci China Inf Sci, 2023, 66: 130301

    work page 2023

  8. [7]

    Viewing channel as sequence rather than image: A 2-D Seq2Seq approach for efficient MIMO-OFDM CSI feedback

    Chen Z R, Zhang Z Y, Xiao Z R, et al. Viewing channel as sequence rather than image: A 2-D Seq2Seq approach for efficient MIMO-OFDM CSI feedback. IEEE Trans Wireless Commun, 2023, 22: 7393-7407

    work page 2023

Show all 61 references
  1. [8]

    C-GRBFnet: A physics-inspired generative deep neural network for channel representation and prediction

    Xiao Z R, Zhang Z Y, Huang C W, et al. C-GRBFnet: A physics-inspired generative deep neural network for channel representation and prediction. IEEE J Sel Areas Commun, 2022. 40: 2282-2299

    2022

  2. [9]

    From data-driven learning to physics-inspired inferring: A novel mobile MIMO channel prediction scheme based on neural ODE

    Xiao Z R, Zhang Z Y, Chen Z R, et al. From data-driven learning to physics-inspired inferring: A novel mobile MIMO channel prediction scheme based on neural ODE. IEEE Trans Wireless Commun, 2024, 23: 7186-7199

    2024

  3. [10]

    Deep learning-based multi-user positioning in wireless FDMA cellular networks

    Chen Z R, Zhang Z Y, Xiao Z R, et al. Deep learning-based multi-user positioning in wireless FDMA cellular networks. IEEE J Sel Areas Commun, 2023, 41: 3848-3862

    2023

  4. [11]

    Channel mapping based on interleaved learning with complex-domain MLP-Mixer

    Chen Z R, Zhang Z Y, Yang Z H, et al. Channel mapping based on interleaved learning with complex-domain MLP-Mixer. IEEE Wireless Commun Lett, 2024, 13: 1369-1373

    2024

  5. [12]

    Deep CSI compression for dual-polarized massive MIMO channels with disentangled representation learning

    Fan S H, Xu W, Xie R J, et al. Deep CSI compression for dual-polarized massive MIMO channels with disentangled representation learning. IEEE Trans Commun, 2024, 72: 5564-5580

    2024

  6. [13]

    Channel deduction: A new learning framework to acquire channel from outdated samples and coarse estimate

    Chen Z R, Zhang Z Y, Yang Z H, et al. Channel deduction: A new learning framework to acquire channel from outdated samples and coarse estimate. 2024. ArXiv:2403.19409

    2024 arXiv

  7. [14]

    Deep learning-based CSI feedback for beamforming in single- and multi-cell massive MIMO systems

    Guo J, Wen C K, Jin S. Deep learning-based CSI feedback for beamforming in single- and multi-cell massive MIMO systems. IEEE J Sel Areas Commun, 2021, 39: 1872-1884

    2021

  8. [15]

    On the opportunities and risks of foundation models

    Bommasani R, Hudson D A, Adeli E, et al. On the opportunities and risks of foundation models. 2021. ArXiv:2108.07258

    2021 arXiv

  9. [16]

    Deepseek-v2: A strong, economical, and efficient mixture-of-experts language model

    Liu A X, Feng B, Wang B, et al. Deepseek-v2: A strong, economical, and efficient mixture-of-experts language model. 2024. ArXiv:2405.04434

    2024 arXiv

  10. [17]

    Deepseek-v3 technical report

    Liu A X, Feng B, Xue B, et al. Deepseek-v3 technical report. 2024. ArXiv:2412.19437

    2024 arXiv

  11. [18]

    Qwen2.5 Technical Report

    Yang A N, Yang B S, Zhang B C, et al. Qwen2.5 Technical Report. 2024. ArXiv:2412.15115

    2024 arXiv

  12. [20]

    Observations on LLMs for telecom domain: Capabilities and limitations

    Soman S, Ranjani H G. Observations on LLMs for telecom domain: Capabilities and limitations. In: Proceedings of the Third International Conference on AI-ML Systems, 2023. 1-5

    2023

  13. [21]

    Large generative AI models for telecom: The next big thing? IEEE Commun Mag, 2024, 62: 84-90

    Bariah L, Zhao Q Y, Zou H, et al. Large generative AI models for telecom: The next big thing? IEEE Commun Mag, 2024, 62: 84-90

    2024

  14. [22]

    Understanding telecom language through large language models

    Bariah L, Zou H, Zhao Q Y, et al. Understanding telecom language through large language models. In: Proceedings of IEEE Global Communications Conference (GLOBECOM), 2023. 6542-6547

    2023

  15. [23]

    Linguistic intelligence in large language models for telecommunications

    Ahmed T, Piovesan N, De Domenico A, et al. Linguistic intelligence in large language models for telecommunications. 2024. ArXiv:2402.15818

    2024 arXiv

  16. [24]

    TKG : Telecom knowledge governance framework for LLM application, 2023

    Cai H R, Wu S J. TKG : Telecom knowledge governance framework for LLM application, 2023. doi: 10.21203/rs.3.rs-3252192/v1

    2023 doi

  17. [25]

    A primer on generative AI for telecom: From theory to practice

    Lin X Q, Kundu L, Dick C, et al. A primer on generative AI for telecom: From theory to practice. 2024. ArXiv:2408.09031

    2024 arXiv

  18. [26]

    Large language models for telecom: Forthcoming impact on the industry

    Maatouk A, Piovesan N, Ayed F, et al. Large language models for telecom: Forthcoming impact on the industry. IEEE Commun Mag, 2024, 63: 62-68

    2024

  19. [27]

    Using large language models to understand telecom standards

    Karapantelakis A, Thakur M, Nikou A, et al. Using large language models to understand telecom standards. In: Proceedings of IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN), 2024. 440-446

    2024

  20. [28]

    Large language model (LLM)-enabled in-context learning for wireless network optimization: A case study of power control

    Zhou H, Hu C M, Yuan D, et al. Large language model (LLM)-enabled in-context learning for wireless network optimization: A case study of power control. 2024. ArXiv:2408.00214

    2024 arXiv

  21. [29]

    Large language models in wireless application design: In-context learning-enhanced automatic network intrusion detection

    Zhang H, Sediq A B, Afana A, et al. Large language models in wireless application design: In-context learning-enhanced automatic network intrusion detection. 2024. ArXiv:2405.11002

    2024 arXiv

  22. [30]

    LLM-empowered resource allocation in wireless communications systems

    Lee W, Park J. LLM-empowered resource allocation in wireless communications systems. 2024. ArXiv:2408.02944

    2024 arXiv

  23. [31]

    Leveraging large language models for wireless symbol detection via in-context learning

    Abbas M, Kar K, Chen T Y. Leveraging large language models for wireless symbol detection via in-context learning. 2024. ArXiv:2409.00124

    2024 arXiv

  24. [32]

    WirelessLLM: Empowering large language models towards wireless intelligence

    Shao J W, Tong J W, Wu Q, et al. WirelessLLM: Empowering large language models towards wireless intelligence. 2024. ArXiv:2405.17053

    2024 arXiv

  25. [33]

    Mobile-LLaMA: Instruction fine-tuning open-source LLM for network analysis in 5G networks

    Kan K B, Mun H, Cao G H, et al. Mobile-LLaMA: Instruction fine-tuning open-source LLM for network analysis in 5G networks. IEEE Network, 2024, 38: 76-83

    2024

  26. [34]

    TelecomGPT: A framework to build telecom-specfic large language models

    Zou H, Zhao Q Y, Tian Y, et al. TelecomGPT: A framework to build telecom-specfic large language models. 2024. ArXiv:2407.09424

    2024 arXiv

  27. [35]

    Large language model (LLM) for telecommunications: A comprehensive survey on principles, key techniques, and opportunities

    Zhou H, Hu C M, Yuan Y, et al. Large language model (LLM) for telecommunications: A comprehensive survey on principles, key techniques, and opportunities. 2024. ArXiv:2405.10825

    2024 arXiv

  28. [36]

    Large language models (LLMs) assisted wireless network deployment in urban settings

    Sevim N, Ibrahim M, Ekin S. Large language models (LLMs) assisted wireless network deployment in urban settings. 2024. ArXiv:2405.13356

    2024 arXiv

  29. [37]

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

    Bornea A L, Ayed F, De Domenico A, et al. Telco-RAG: Navigating the challenges of retrieval-augmented language models for telecommunications. 2024. ArXiv:2404.15939

    2024 arXiv

  30. [38]

    TelecomRAG: Taming telecom standards with retrieval augmented generation and LLMs

    Yilma G M, Ayala-Romero J A, Garcia-Saavedra A, et al. TelecomRAG: Taming telecom standards with retrieval augmented generation and LLMs. 2024. ArXiv:2406.07053

    2024 arXiv

  31. [39]

    Telecom language models: Must they be large? 2024

    Piovesan N, De Domenico A, Ayed F. Telecom language models: Must they be large? 2024. ArXiv:2403.04666

    2024 arXiv

  32. [40]

    Unlocking telecom domain knowledge using LLMs

    Roychowdhury S, Jain N, Soman S. Unlocking telecom domain knowledge using LLMs . In: Proceedings of IEEE International Conference on Communication Systems and Networks (COMSNETS), 2024. 267-269

    2024

  33. [41]

    Design of a large language model for improving customer service in telecom operators

    Ma X L, Zhao R Q, Liu Y, et al. Design of a large language model for improving customer service in telecom operators. Electron Lett, 2024, 60: 13218

    2024

  34. [42]

    SPEC5G : A dataset for 5G cellular network protocol analysis

    Karim I, Mubasshir K S, Rahman M M, et al. SPEC5G : A dataset for 5G cellular network protocol analysis. 2023. ArXiv:2301.09201

    2023 arXiv

  35. [43]

    Teleqna : A benchmark dataset to assess large language models telecommunications knowledge

    Maatouk A, Ayed F, Piovesan N, et al. Teleqna : A benchmark dataset to assess large language models telecommunications knowledge. 2023. ArXiv:2310.15051

    2023 arXiv

  36. [44]

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

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

    2024 arXiv

  37. [45]

    Tele-LLMs : A series of specialized large language models for telecommunications

    Maatouk A, Ampudia KC, Ying R, et al. Tele-LLMs : A series of specialized large language models for telecommunications. 2024. ArXiv:2409.05314

    2024 arXiv

  38. [46]

    WirelessAgent: Large language model agents for intelligent wireless networks

    Tong J W, Shao J W, Wu Q, et al. WirelessAgent: Large language model agents for intelligent wireless networks. 2024. ArXiv:2409.07964

    2024 arXiv

  39. [47]

    LLM agents as 6G orchestrator: A paradigm for task-oriented physical-layer automation

    Xiao Z R, Ye C H, Hu Y B, et al. LLM agents as 6G orchestrator: A paradigm for task-oriented physical-layer automation. 2024. ArXiv:2410.03688

    2024 arXiv

  40. [48]

    LLMind: Orchestrating AI and IoT with LLM for complex task execution

    Cui H W, Du Y Y, Yang Q, et al. LLMind: Orchestrating AI and IoT with LLM for complex task execution. IEEE Commun Mag. 2024. 1-7

    2024

  41. [49]

    When large language model agents meet 6G networks: Perception, grounding, and alignment

    Xu M R, Niyato D, Kang J W, et al. When large language model agents meet 6G networks: Perception, grounding, and alignment. IEEE Wireless Commun, 2024, 31: 63-71

    2024

  42. [50]

    Large language model enhanced multi-agent systems for 6G communications

    Jiang F B, Peng Y B, Dong L, et al. Large language model enhanced multi-agent systems for 6G communications. IEEE Wireless Commun, 2024, 31: 48-55

    2024

  43. [51]

    Deep learning for massive MIMO CSI feedback

    Wen C K, Shih W T, Jin S. Deep learning for massive MIMO CSI feedback. IEEE Wireless Commun Lett, 2018, 7: 748-751

    2018

  44. [52]

    Fingerprint-based localization for massive MIMO-OFDM system With deep convolutional neural networks

    Sun X Y, Wu C, Gao X Q, et al. Fingerprint-based localization for massive MIMO-OFDM system With deep convolutional neural networks. IEEE Trans Veh Technol, 2019, 68: 10846-10857

    2019

  45. [53]

    Federated learning with unsourced random access

    Tian Y Q, Che J Z, Zhang Z Y, et al. Federated learning with unsourced random access. In: Proceedings of 2023 IEEE 97th Vehicular Technology Conference (VTC2023-Spring), 2023. 1-5

    2023

  46. [54]

    CSI-GPT : Integrating generative pre-trained transformer with federated-tuning to acquire downlink massive MIMO channels

    Zeng Y, Qiao L, Gao Z, et al. CSI-GPT : Integrating generative pre-trained transformer with federated-tuning to acquire downlink massive MIMO channels. 2024. ArXiv:2406.03438

    2024 arXiv

  47. [55]

    Csi-LLM: A novel downlink channel prediction method aligned with LLM pre-training

    Fan S L, Liu Z Y, Gu X Y, et al. Csi-LLM: A novel downlink channel prediction method aligned with LLM pre-training. 2024 ArXiv:2409.00005

    2024 arXiv

  48. [56]

    LLM4CP : Adapting large language models for channel prediction

    Liu B X, Liu X Y, Gao S J, et al. LLM4CP : Adapting large language models for channel prediction. 2024. ArXiv:2406.14440

    2024 arXiv

  49. [57]

    Assessing air-interface dataset similarity and diversity for AI-enabled wireless communications

    Tang H N, Yang L S, Zhou R, et al. Assessing air-interface dataset similarity and diversity for AI-enabled wireless communications. In: Proceedings of IEEE International Conference on Communications Workshops (ICC Workshops), 2024. 1623-1628

    2024

  50. [58]

    VBIM-Net: Variational Born Iterative Network for Inverse Scattering Problems

    Xing Z Q, Zhang Z Y, Chen Z R, et al. VBIM-Net: Variational Born Iterative Network for Inverse Scattering Problems. IEEE Trans Geosci Remote Sensing, 2025, 63: 1-16

    2025

  51. [59]

    Meta-Material Sensor-Based Internet of Things for Environmental Monitoring by Deep Learning: Design, Deployment, and Implementation

    Liu X, Hu J Z, Zhang H L, et al. Meta-Material Sensor-Based Internet of Things for Environmental Monitoring by Deep Learning: Design, Deployment, and Implementation. IEEE Trans Wireless Commun, 2023, 22: 2462-2476

    2023

  52. [60]

    Stepsize-Adaptive SAMP Algorithm for Fast mmWave Radar Imaging

    Zhang C Z, Zhang Z Y, Chen Z R, et al. Stepsize-Adaptive SAMP Algorithm for Fast mmWave Radar Imaging. In: Proceedings of 2024 IEEE Wireless Communications and Networking Conference (WCNC), 2024. 1-6

    2024

  53. [61]

    Wireless Federated Learning Over Resource-Constrained Networks: Digital Versus Analog Transmissions

    Yao J C, Xu W, Yang Z H, et al. Wireless Federated Learning Over Resource-Constrained Networks: Digital Versus Analog Transmissions. IEEE Trans Wireless Commun, 2024. 14020-14036

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.