Pith. sign in

REVIEW 4 major objections 5 minor 40 references

ModFus-DM: Explore the Representation in Modulated Signal Diffusion Generated Models

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Diffusion models trained to denoise unlabeled IQ signals produce modulation representations that rival fully supervised classifiers using only a handful of labeled examples per modulation type.

desk verdict A coherent diffusion-representation approach to AMC with a serious verification problem in its key ablation table, so the headline accuracy claims are not yet trustworthy. read the letter →

arxiv 2508.01719 v1 pith:NP5ATKLL submitted 2025-08-03 eess.SP

Haoyue Tan , Yu Li , Zhenxi Zhang , Xiaoran Shi , Feng Zhou This is my paper · ORCID
classification eess.SP
keywords automaticmodulationclassificationdiffusionmodelsself-supervisedrepresentationlearninglimited-labelU-Netfeaturesrecognitionvariable-lengthsignals
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

This paper argues that a diffusion model trained purely on unlabeled radio signals can learn a representation of modulation type that is reusable and highly label-efficient. The proposed framework, ModFus-DM, first trains a noisy-signal reconstruction model (MSDGM) on raw IQ data, then freezes it and trains only a lightweight fusion module and classifier on a handful of labeled signals per class. The paper reports that with 10 labeled signals per modulation type it reaches over 88% accuracy on a 24-class recognition task at high SNR, and that the same frozen features generalize across datasets, signal lengths, and fading channels. If true, this would make generative self-supervision a practical source of modulation representations where annotation is scarce.

What carries the argument

The modulated signal diffusion generation model (MSDGM): a diffusion model over IQ signals in which noise is added over T steps and a U-Net predicts the noise, trained with the standard denoising loss. The diffusion-aware feature fusion module (DAFFus) then takes the signals st at a small diffusion step t, extracts features from the U-Net's upsampling blocks b5-b8, concatenates them along the channel dimension, and projects them through a linear layer into a compact feature vector used for classification. The key operation is the adaptive fusion of these multi-scale features, which the paper shows outperforms any single block's features and both all-block and downsampling-only fusion.

What would settle it

Collect naturally recorded (uncropped) signals at lengths 64, 128, 256, 512, and 768 with known modulation labels, and evaluate the frozen MSDGM plus DAFFus trained on length-1024 data; if accuracy on the naturally short signals drops far below the random-crop results, or if the model fails on a length not seen in training, the temporal-flexibility claim would be refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that intermediate features of a U-Net trained to denoise modulated signals carry discriminative modulation information, and that adaptively fusing the upsampling-block features yields a representation that separates modulation types better than any single layer or than all-block or downsampling-only fusion. The generative model, called MSDGM, is trained on raw unlabeled IQ signals with a standard noise-prediction loss; afterwards it is frozen and only the DAFFus fusion module and classifier are trained on as few as two to twenty labeled signals per class. Experiments on RML2016.10A, RML2016.10B, RML2018.01A, and RML2022 are reported as showing consistent gains over supervised and self-supervised baselines in limited-label settings, cross-dataset transfer, variable-length inputs, and fading channels.

Load-bearing premise

The claim that a single trained model recognizes signals of any length assumes that randomly cropping long signals into short ones produces the same conditions as naturally captured short signals, and that a U-Net trained on one length can process any length without architectural changes.

Editorial extensions

If this is right

  • Radio receivers could build recognition systems from unlabeled over-the-air captures, reducing annotation cost by orders of magnitude in spectrum monitoring and cognitive radio.
  • A single frozen diffusion feature extractor could be reused across datasets, SNRs, and channel conditions, so new deployments would only need to train a small fusion head.
  • Robustness to fading channels and colored noise suggests the learned representation captures signal structure that survives realistic propagation effects.
  • Length-flexible inference without architectural changes would allow models to process naturally variable-length transmissions instead of fixed-size windows.

Reading between the lines

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

  • The frozen representation should be label-set agnostic, so one could test whether DAFFus trained on one set of modulation types transfers to newly added types without retraining the diffusion model.
  • Since DAFFus uses features at a fixed denoising step t=1, a natural extension is to fuse features across several diffusion steps, which may further improve robustness to noise.
  • The superiority of upsampling-block features is demonstrated only for this U-Net and IQ-signal domain; comparing DAFFus on a masked-autoencoder backbone would separate the effect of diffusion from the effect of generative self-supervision in general.
  • If the representation is truly length-invariant, it should support zero-shot classification of modulation types that were absent from the labeled subset, testable with a nearest-prototype classifier over the frozen features.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes ModFus-DM, a two-stage framework for automatic modulation classification. In the first stage, an unsupervised modulated-signal diffusion model (MSDGM) is trained on raw IQ signals to learn representations without labels or proxy tasks. In the second stage, the diffusion U-Net is frozen and a diffusion-aware feature fusion module (DAFFus) adaptively aggregates features from several U-Net blocks, followed by a linear classifier trained with a limited number of labeled signals per modulation type. The authors report extensive experiments on RML2016.10A, RML2016.10B, RML2018.01A, and RML2022, claiming consistent improvements over supervised, self-supervised, and semi-supervised baselines under limited-label settings, distribution shift, variable-length inputs, channel fading, and colored noise. The headline claim is over 88.27% accuracy on RML2018.01A with only 10 labeled signals per type.

Significance. If the empirical claims hold, the paper would make a useful contribution by showing that frozen diffusion-model features, trained without labels, provide a reuseable and label-efficient representation for radio modulation signals. The paper has notable strengths: it evaluates on four benchmarks, includes several challenging settings (limited labels, cross-dataset transfer, variable lengths, fading channels, colored noise), and compares against a broad set of baselines. The idea of using the diffusion process as an unsupervised representation learner for IQ signals is timely and plausible. However, the significance is currently gated by verification concerns in the central experimental tables, the lack of uncertainty quantification, and the use of test-set-based selection for key hyperparameters.

major comments (4)
  1. [Section V.B.2, Table II] The T=100 row is internally inconsistent. The five per-SNR accuracies listed are 90.65, 92.93, 93.26, 93.96, and 92.48; their arithmetic mean is 92.66, not the reported 89.29. Every other row in the same table has a Mean exactly equal to the arithmetic mean of its five entries, so the intended definition is unambiguous. This matters because T=100 is the configuration used in all subsequent experiments, including the headline 88.27% result in Table III. As the paper stands, a load-bearing summary statistic cannot be reproduced from the table itself. The authors should provide corrected numbers, raw per-SNR logs, or code so that this discrepancy can be resolved.
  2. [Section V.A and all experimental figures] The paper states that ten Monte Carlo experiments are run and the average accuracy is reported, but no error bars, standard deviations, or confidence intervals appear anywhere in the paper. Many reported advantages are large (e.g., Fig. 6, where ModFus-DM surpasses the second-best method by roughly 40–60 percentage points at N=2), but without variance information the reader cannot judge whether smaller differences, such as the 1.12% gap between DAFFus and Fusion-all in Fig. 4, are meaningful. I request that the authors report the spread of the ten runs, at least for the main tables and for the key limited-label comparisons.
  3. [Section V.B.1, Fig. 4, and downstream experiments] The diffusion step t=1 and the DAFFus fusion set (blocks b5–b8) are selected based on accuracy on the same benchmark test sets that are later used for the headline claims. The text says 'we fix t = 1 for all subsequent experiments' after observing that small t gives good results on RML2016.10A, RML2016.10B, RML2018, and RML2022. This is selection on test data. Since the central claim is an empirical performance claim, the reported numbers are optimistically biased unless the selection is accounted for, for example by reporting a validation split or by demonstrating that the chosen configuration is best across datasets in a pre-registered way. Please clarify the selection protocol or re-evaluate with a proper validation split.
  4. [Section V.F, Table IV, and Fig. 9] The temporal-flexibility claims are based on signals of lengths 64, 128, 256, 512, and 768 that are randomly cropped from full-length 1024 samples, as stated in the Table IV note. This is a proxy for natural variable-length signals, and the paper does not validate it against naturally captured variable-length data or compare baseline methods under the same crop protocol. Moreover, the U-Net is trained on length-1024 inputs and then applied to other lengths without any discussion of architectural changes or positional-encoding adaptation. If the crop proxy or the length-flexibility assumption fails, the temporal-flexibility conclusion in Section V.F would collapse. The authors should either add a baseline comparison under the same crops or temper the claim to what the experiment actually supports.
minor comments (5)
  1. [Abstract and Table III] The abstract says 'over 88.27% accuracy in 24-type recognition tasks at SNR ≥ 12dB', but 88.27% appears to be the single 12dB N=10 entry in Table III. Please clarify whether the claim refers to a single SNR point or to an average over SNR ≥ 12 dB.
  2. [Fig. 2 caption] The caption refers to 'MSGDM' while the text uses 'MSDGM'. Please correct the acronym for consistency.
  3. [Section V.G] The text contains the typo 'Rayleighigh' and later 'Rayleighigh fading'. Please fix these typos.
  4. [Eq. (13)] The definition of AP as 'pooling the final dimension to 1' is ambiguous. Please specify whether the pooling is over the time dimension, the channel dimension, or both, and how the pooled features are shaped.
  5. [General presentation] The paper repeatedly describes ModFus-DM as a 'foundational model' and the 'first' such model for AMC. Given that the method is a frozen diffusion feature extractor plus a learned fusion head, the 'foundational model' claim is stronger than the evidence presented. Please soften or justify this terminology.

Circularity Check

0 steps flagged · score 0.0 of 10

No definitional circularity: the benchmark claims rest on held-out evaluation of a frozen unsupervised representation plus a small labeled classifier, not on a fitted quantity renamed as a prediction.

full rationale

ModFus-DM's central claim is an empirical benchmark claim. The representation model is trained on unlabeled signals with the standard noise-prediction objective LM = E||epsilon - delta_theta(s_t,t)||^2, which contains no label term; it is then frozen. DAFFus and the classifier are trained on N labeled signals per class per SNR with the cross-entropy Eq. 17 and evaluated on separate test partitions of RML2016.10A/B, RML2018.01A, and RML2022. The reported 88.27% or Fig. 7 accuracies are held-out classification results, not quantities defined by Eqs. 12-17 by construction. There is no fitted parameter, fusion weight, or diffusion step that is algebraically forced to equal the reported accuracy. The self-citations in the paper (refs. [5], [9], [31] and related items by overlapping authors) are background and related-work descriptions; none is used as a load-bearing uniqueness theorem or as the justification for the main result, and ref. [31] is cited only as a prior contrastive unsupervised AMC method. The selection of t=1 and of the b5-b8 fusion set is made from Fig. 4, an ablation on the same evaluation protocol; this is hyperparameter selection on the test protocol rather than a definitional reduction, and it does not turn the final accuracy into an input of the derivation. One verification concern is explicitly outside the circularity pass: Table II's T=100 row is internally inconsistent (90.65, 92.93, 93.26, 93.96, 92.48 average to 92.66, not the printed 89.29), and T=100 is used in all downstream tables; that is an arithmetic/reproducibility risk, not a self-derivation. Consequently, no step in the paper exhibits the equation-level or renamed-fit pattern required for a circularity finding.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard diffusion math, the representation-learning premise from prior image-domain work, and several design choices (t=1, T=100, b5-b8, d=128) selected by ablation on the evaluation benchmarks. No new physical entities or forces are introduced. The variable-length generalization additionally assumes random crops are valid stand-ins for realistic variable-length signals.

free parameters (5)
  • Diffusion step t used for feature extraction = t=1
    Selected because the ablation in Fig. 4 shows highest or near-highest accuracy at small t, then fixed for all downstream experiments.
  • Total number of diffusion steps T = T=100
    Chosen by ablation on RML2016.10A (Table II) as the point where accuracy saturates; the table itself contains an internal inconsistency.
  • Multiplicative constant in diffusion schedule nu_t proportional to 1/t = not specified
    The schedule is only defined up to a constant, so the actual noise level at each step is not reproducible from the paper.
  • DAFFus fusion layer set = upsampling blocks b5-b8, L=8 total
    Chosen from the Fig. 4 ablation over blocks and fusion variants; not derived from theory.
  • DAFFus feature dimension d = d=128
    Hand-set hyperparameter; no sensitivity analysis is provided.
assumptions (4)
  • standard math The reverse denoising network can approximate the true Gaussian posterior at each step (Eqs. 7-10).
    Standard diffusion-model posterior from the Gaussian product theorem; the paper invokes it without proof.
  • domain assumption Intermediate U-Net layers trained for denoising encode modulation-discriminative semantics.
    Adopted from prior representation-learning results [19][20]; Section IV.C uses it to justify extracting features from blocks b1-b8.
  • domain assumption Randomly cropping length-1024 signals produces valid variable-length test signals that preserve modulation identity.
    Section V.F and Table IV use crops of lengths 64-768 as a variable-length benchmark without validating that cropping preserves usable modulation information or matches realistic variable-length acquisition.
  • domain assumption Baseline models in the comparisons are trained under the same limited-label and SNR protocol with equivalent hyperparameter tuning.
    Section V.A and Figs. 6-7 state the protocol but do not provide baseline implementation details, seeds, or tuning budgets.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ModFus-DM: Explore the Representation in Modulated Signal Diffusion Generated Models." pith.science (2026). https://pith.science/paper/NP5ATKLL

@misc{pith2026250801719,
  author       = {Pith},
  title        = {Pith review of: ModFus-DM: Explore the Representation in Modulated Signal Diffusion Generated Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NP5ATKLL}},
  note         = {Machine review of arXiv:2508.01719}
}
abstract

Automatic modulation classification (AMC) is essential for wireless communication systems in both military and civilian applications. However, existing deep learning-based AMC methods often require large labeled signals and struggle with non-fixed signal lengths, distribution shifts, and limited labeled signals. To address these challenges, we propose a modulation-driven feature fusion via diffusion model (ModFus-DM), a novel unsupervised AMC framework that leverages the generative capacity of diffusion models for robust modulation representation learning. We design a modulated signal diffusion generation model (MSDGM) to implicitly capture structural and semantic information through a progressive denoising process. Additionally, we propose the diffusion-aware feature fusion (DAFFus) module, which adaptively aggregates multi-scale diffusion features to enhance discriminative representation. Extensive experiments on RML2016.10A, RML2016.10B, RML2018.01A and RML2022 datasets demonstrate that ModFus-DM significantly outperforms existing methods in various challenging scenarios, such as limited-label settings, distribution shifts, variable-length signal recognition and channel fading scenarios. Notably, ModFus-DM achieves over 88.27% accuracy in 24-type recognition tasks at SNR $\geq $ 12dB with only 10 labeled signals per type.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 36 canonical work pages

  1. [1]

    Lightweight automatic modulation classification via progres- sive differentiable architecture search,

    X. Zhang, X. Chen, Y . Wang, G. Gui, B. Adebisi, H. Sari, and F. Adachi, “Lightweight automatic modulation classification via progres- sive differentiable architecture search,” IEEE Transactions on Cognitive Communications and Networking , vol. 9, no. 6, pp. 1519–1530, 2023

    work page 2023

  2. [2]

    Novel automatic modulation classification using cumulant features for communications via multipath channels,

    H.-C. Wu, M. Saquib, and Z. Yun, “Novel automatic modulation classification using cumulant features for communications via multipath channels,” IEEE Transactions on Wireless Communications, vol. 7, no. 8, pp. 3098–3105, 2008

    work page 2008

  3. [3]

    Psrnet: Few-shot automatic modulation classification under potential domain differences,

    H. Xing, S. Wang, J. Wang, L. Mei, Y . Xu, H. Zhou, H. Xu, and L. Jiao, “Psrnet: Few-shot automatic modulation classification under potential domain differences,” IEEE Transactions on Wireless Communications , vol. 24, no. 1, pp. 371–384, 2025

    work page 2025

  4. [4]

    Shared spectrum monitoring using deep learning,

    F. A. Bhatti, M. J. Khan, A. Selim, and F. Paisana, “Shared spectrum monitoring using deep learning,” IEEE Transactions on Cognitive Com- munications and Networking , vol. 7, no. 4, pp. 1171–1185, 2021

    work page 2021

  5. [5]

    Pass-net: A pseudo classes and stochastic classifiers based network for few-shot class-incremental automatic modulation classification,

    H. Tan, Z. Zhang, Y . Li, X. Shi, L. Wang, X. Yang, and F. Zhou, “Pass-net: A pseudo classes and stochastic classifiers based network for few-shot class-incremental automatic modulation classification,” IEEE Transactions on Wireless Communications , vol. 23, no. 12, pp. 17 987– 18 003, 2024

    work page 2024

  6. [6]

    Ofdm receiver design with learning-driven automatic modulation recognition,

    L. P. Qian, C. Wang, Q. Wang, M. Wu, Y . Wu, and X. Yang, “Ofdm receiver design with learning-driven automatic modulation recognition,” IEEE Transactions on Cognitive Communications and Networking , vol. 10, no. 2, pp. 429–441, 2024

    work page 2024

  7. [7]

    One2threenet: An automatic microscale-based modulation recognition method for underwater acous- tic communication systems,

    J. Wang, Z. Huang, W. Shi, and S. Mao, “One2threenet: An automatic microscale-based modulation recognition method for underwater acous- tic communication systems,” IEEE Transactions on Wireless Communi- cations, vol. 23, no. 8, pp. 10 287–10 300, 2024

    work page 2024

  8. [8]

    Afd-il: A long- term incremental learning approach with adaptive feature distillation for specific emitter identification,

    L. Sun, R. Xue, H. Zha, Q. Tian, W. Wang, and Y . Lin, “Afd-il: A long- term incremental learning approach with adaptive feature distillation for specific emitter identification,” IEEE Transactions on Cognitive Communications and Networking , pp. 1–1, 2025

    work page 2025

Show all 40 references
  1. [9]

    Multi-scale feature fusion and distribution similarity network for few-shot automatic modulation classification,

    H. Tan, Z. Zhang, Y . Li, X. Shi, and F. Zhou, “Multi-scale feature fusion and distribution similarity network for few-shot automatic modulation classification,” IEEE Signal Processing Letters , pp. 1–5, 2024

    2024

  2. [10]

    A survey of deep transfer learning in automatic modulation classification,

    X. Wang, Y . Zhao, and Z. Huang, “A survey of deep transfer learning in automatic modulation classification,” IEEE Transactions on Cognitive Communications and Networking , pp. 1–1, 2025

    2025

  3. [11]

    A transformer-based contrastive semi-supervised learning framework for automatic modula- tion recognition,

    W. Kong, X. Jiao, Y . Xu, B. Zhang, and Q. Yang, “A transformer-based contrastive semi-supervised learning framework for automatic modula- tion recognition,” IEEE Transactions on Cognitive Communications and Networking, vol. 9, no. 4, pp. 950–962, Aug. 2023

    2023

  4. [12]

    MCLHN: Toward automatic modulation classification via masked contrastive learning with hard negatives,

    C. Xiao, S. Yang, Z. Feng, and L. Jiao, “MCLHN: Toward automatic modulation classification via masked contrastive learning with hard negatives,” IEEE Transactions on Wireless Communications , vol. 23, no. 10, pp. 14 304–14 319, Oct. 2024

    2024

  5. [13]

    Unsupervised modulation recognition method based on multi-domain representation contrastive learning,

    Y . Li, X. Shi, X. Yang, and F. Zhou, “Unsupervised modulation recognition method based on multi-domain representation contrastive learning,” in 2023 IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC) , Nov. 2023, pp. 1–6

    2023

  6. [14]

    Auto-encoding variational bayes,

    M. Diederik P. Kingma, “Auto-encoding variational bayes,” in Pro- ceedings of the International Conference on Learning Representations (ICLR), 2014

    2014

  7. [15]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” in Advances in Neural Information Processing Systems , vol. 27, 2014

    2014

  8. [16]

    Data augmentation aided automatic modulation recognition using diffusion model,

    J. Chen, C. Zhao, X. Huang, and Z. Wu, “Data augmentation aided automatic modulation recognition using diffusion model,” in 2024 IEEE Wireless Communications and Networking Conference (WCNC) , Apr. 2024, pp. 1–6

    2024

  9. [17]

    Diffusion model empowered data augmentation for automatic modulation recognition,

    M. Li, P. Wang, Y . Dong, and Z. Wang, “Diffusion model empowered data augmentation for automatic modulation recognition,” IEEE Wireless Communications Letters , pp. 1–1, 2025

    2025

  10. [18]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” Oct. 2022, arXiv:2010.02502. [Online]. Available: http: //arxiv.org/abs/2010.02502

    2022 arXiv

  11. [19]

    Denoising diffusion autoencoders are unified self-supervised learners,

    W. Xiang, H. Yang, D. Huang, and Y . Wang, “Denoising diffusion autoencoders are unified self-supervised learners,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 15 802–15 812

    2023

  12. [20]

    Diffusion model as representation learner,

    X. Yang and X. Wang, “Diffusion model as representation learner,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 18 938–18 949

    2023

  13. [21]

    A contrastive learner for automatic modu- lation classification,

    M. Du, J. Pan, and D. Bi, “A contrastive learner for automatic modu- lation classification,” IEEE Transactions on Wireless Communications , vol. 24, no. 4, pp. 3575–3589, 2025

    2025

  14. [22]

    Gaf-mae: A self- supervised automatic modulation classification method based on gramian angular field and masked autoencoder,

    Y . Shi, H. Xu, Y . Zhang, Z. Qi, and D. Wang, “Gaf-mae: A self- supervised automatic modulation classification method based on gramian angular field and masked autoencoder,” IEEE Transactions on Cognitive Communications and Networking , vol. 10, no. 1, pp. 94–106, 2024

    2024

  15. [23]

    Over-the-air deep learnig based radio signal classification,

    T. J. O’Shea, T. Roy, and T. C. Clancy, “Over-the-air deep learnig based radio signal classification,” IEEE Journal of Selected Topics in Signal Processing, vol. 12, no. 1, pp. 168–179, Feb. 2018

    2018

  16. [24]

    Deep neural network architectures for modulation classification,

    X. Liu, D. Yang, and A. E. Gamal, “Deep neural network architectures for modulation classification,” in 2017 51st Asilomar Conference on Signals, Systems, and Computers , Oct. 2017, pp. 915–919

    2017

  17. [25]

    Real-time radio technology and modulation classification via an lstm auto-encoder,

    Z. Ke and H. Vikalo, “Real-time radio technology and modulation classification via an lstm auto-encoder,” IEEE Transactions on Wireless Communications, vol. 21, no. 1, pp. 370–382, Jan. 2022

    2022

  18. [26]

    A spatiotemporal multi-channel learning framework for automatic modulation recognition,

    J. Xu, C. Luo, G. Parr, and Y . Luo, “A spatiotemporal multi-channel learning framework for automatic modulation recognition,” IEEE Wire- less Communications Letters , vol. 9, no. 10, pp. 1629–1632, Oct. 2020

    2020

  19. [27]

    Automatic modulation classification based on complex-valued convolutional neural network and semi-supervised learning,

    C. Liu, Z. Cai, and B. Zhang, “Automatic modulation classification based on complex-valued convolutional neural network and semi-supervised learning,” in 2023 IEEE 23rd International Conference on Communica- tion Technology (ICCT) , Wuxi, China, Oct. 2023, pp. 241–246

    2023

  20. [28]

    A transformer-based CTDNN structure for automatic modulation recognition,

    W. Kong, Q. Yang, X. Jiao, Y . Niu, and G. Ji, “A transformer-based CTDNN structure for automatic modulation recognition,” in 2021 7th International Conference on Computer and Communications (ICCC) , Dec. 2021, pp. 159–163

    2021

  21. [29]

    Self-contrastive learning based semi-supervised radio modulation classification,

    D. Liu, P. Wang, T. Wang, and T. Abdelzaher, “Self-contrastive learning based semi-supervised radio modulation classification,” in MILCOM 2021 - 2021 IEEE Military Communications Conference (MILCOM) , Nov. 2021, pp. 777–782

    2021

  22. [30]

    Hybrid-view self- supervised framework for automatic modulation recognition,

    Y . Fu, Y . Ma, Z. Feng, S. Yang, and Y . Wang, “Hybrid-view self- supervised framework for automatic modulation recognition,” IEEE Internet of Things Journal , vol. 12, no. 6, pp. 7360–7375, Mar. 2025

    2025

  23. [31]

    Multi- representation domain attentive contrastive learning based unsupervised automatic modulation recognition,

    Y . Li, X. Shi, H. Tan, Z. Zhang, X. Yang, and F. Zhou, “Multi- representation domain attentive contrastive learning based unsupervised automatic modulation recognition,” Nature Communications , vol. 16, no. 1, p. 5951, 2025

    2025

  24. [32]

    Video diffusion models,

    J. Ho, T. Salimans, A. Gritsenko, W. Chan, M. Norouzi, and D. J. Fleet, “Video diffusion models,” Advances in Neural Information Processing Systems, vol. 35, pp. 8633–8646, 2022

    2022

  25. [33]

    Radio machine learning dataset generation with gnu radio,

    T. J. O’shea and N. West, “Radio machine learning dataset generation with gnu radio,” in Proceedings of the GNU radio conference , vol. 1, no. 1, 2016

    2016

  26. [34]

    Convolutional radio mod- ulation recognition networks,

    T. J. O’Shea, J. Corgan, and T. C. Clancy, “Convolutional radio mod- ulation recognition networks,” in Engineering Applications of Neural Networks, Cham, 2016, pp. 213–226

    2016

  27. [35]

    RML22: Realistic dataset generation for wireless modulation classification,

    V . Sathyanarayanan, P. Gerstoft, and A. E. Gamal, “RML22: Realistic dataset generation for wireless modulation classification,” IEEE Trans- actions on Wireless Communications , vol. 22, no. 11, pp. 7663–7675, Nov. 2023

    2023

  28. [36]

    An efficient deep learning model for automatic modulation recognition based on parameter estimation and transformation,

    F. Zhang, C. Luo, J. Xu, and Y . Luo, “An efficient deep learning model for automatic modulation recognition based on parameter estimation and transformation,” IEEE Communications Letters , vol. 25, no. 10, pp. 3287–3290, Oct. 2021

    2021

  29. [37]

    Open set recognition of communication signal modulation based on deep learning,

    X. Zhang, T. Li, P. Gong, R. Liu, X. Zha, and W. Tang, “Open set recognition of communication signal modulation based on deep learning,” IEEE Communications Letters , vol. 26, no. 7, pp. 1588–1592, Jul. 2022

    2022

  30. [38]

    OSMR: Open- set modulation recognition based on information enhancement,

    Y . Ling, L. Wang, Y . Wang, C. Hou, J. Pan, and N. Ye, “OSMR: Open- set modulation recognition based on information enhancement,” in 2024 International Conference on Ubiquitous Communication (Ucom) , Jul. 2024, pp. 406–410

    2024

  31. [39]

    SSRCNN: A semi-supervised Learning framework for signal recognition,

    Y . Dong, X. Jiang, L. Cheng, and Q. Shi, “SSRCNN: A semi-supervised Learning framework for signal recognition,” IEEE Transactions on Cognitive Communications and Networking , vol. 7, no. 3, pp. 780–789, Sep. 2021

    2021

  32. [40]

    Representation learning with contrastive predictive coding,

    O. V . Aaron van den Oord, Yazhe Li, “Representation learning with contrastive predictive coding,” Jan. 2019, arXiv:1807.03748. [Online]. Available: http://arxiv.org/abs/1807.03748v2

    2019 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.