Federated Learning Assisted Edge Caching Scheme Based on Lightweight Architecture DDPM

Kezhi Wang; Khaled B. Letaief; Nan Cheng; Pingyi Fan; Qiong Wu; Xun Li

REVIEW 4 major objections 5 minor 18 references

Federated Learning Assisted Edge Caching Scheme Based on Lightweight Architecture DDPM

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read According to the paper, a latent-space federated diffusion model predicts popular content for edge caches more accurately than GAN, Thompson sampling, and raw-data diffusion baselines on MovieLens 1M.

desk verdict A plausible FL-plus-lightweight-diffusion caching scheme with real code, but the missing train/test split keeps the headline gains from being fully checkable. read the letter →

arxiv 2506.04593 v3 pith:SRAGFA3J submitted 2025-06-05 cs.NI eess.SP

Xun Li , Qiong Wu , Pingyi Fan , Kezhi Wang , Nan Cheng , Khaled B. Letaief This is my paper

classification cs.NIeess.SP

keywords federatedlearningdenoisingdiffusionprobabilisticmodeledgecachingcontentpopularitypredictionlatentspacelightweightU-Netcachehitpercentageprivacypreservation

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 tries to establish that a small diffusion model can be trained collaboratively across user devices and then used by a base station to predict content popularity accurately enough to outperform existing federated caching schemes. The key move is to compress each user's high-dimensional sparse rating vector into a low-dimensional latent space before diffusion training, which avoids the flat-noise problem that makes diffusion models fail on raw sparse data, and to keep the model small (770K parameters) so resource-limited devices can train it. The authors compare against a federated Wasserstein GAN baseline, Thompson Sampling, a raw-data diffusion model, and an Oracle upper bound on the MovieLens 1M dataset; their scheme reports higher cache hit percentage and lower request delay than all but Oracle. If the result holds, wireless edge caching could get a privacy-preserving generative predictor that runs on ordinary devices rather than server clusters.

What carries the argument

The load-bearing mechanism is a three-stage pipeline: a pre-trained encoder maps each user's 3952-dimensional sparse rating vector into a 16-dimensional latent vector, a 770K-parameter 1D-convolutional U-Net learns the latent distribution via the standard DDPM objective $\mathcal{L}_{t-1}=\mathbb{E}_{t,x_0,\epsilon}\|\epsilon-\epsilon_\theta(\sqrt{\bar{\alpha}_t}x_0+\sqrt{1-\bar{\alpha}_t}\epsilon,t)\|^2$, and a pre-trained decoder reconstructs generated latent samples back into item space. The U-Net is trained under the federated averaging rule of Eq. (9), and content popularity is scored by averaging $U$ decoded samples as in Eq. (10). This pipeline converts diffusion from a compute-heavy server-side model into a device-trainable model while avoiding the flat-noise problem of high-dimensional sparse data.

What would settle it

Re-run the MovieLens 1M experiment with an explicit temporal cutoff: train the encoder, decoder, and federated LDPM only on ratings with timestamps before the cutoff, and measure cache hits only on requests after the cutoff. If the reported advantage over CPPPP and Thompson Sampling disappears or reverses, the central claim is falsified; the paper as written does not report such a split, so this check is necessary to know which case holds.

Watch

Extended reading notes

Core claim

The central claim is that a denoising diffusion probabilistic model can be made practical for federated edge caching if it operates not on raw high-dimensional sparse interaction vectors but on a low-dimensional latent representation learned by a pre-trained encoder. Diffusion models normally need heavy compute, and sparse high-dimensional data makes the noise distribution flat, so the model cannot distinguish signal from noise. The paper's fix is a 770K-parameter 1D-convolution U-Net trained under federated averaging: each user maps their data into latent space with a pre-trained encoder, trains locally, uploads weights, and the base station aggregates them. At prediction time, the global model generates synthetic samples in latent space, a pre-trained decoder reconstructs them into full item-space vectors, and summing these reconstructed samples yields a popularity score per item; the top-N items are cached. The reported result is that this latent-space federated LDPM beats the federated WGAN baseline, Thompson Sampling, and raw-data LDPM in cache hit percentage and request delay on MovieLens 1M.

Load-bearing premise

The reported gains assume the model is trained only on past user requests and evaluated on genuinely future requests, but the paper never documents the train/test split, so the results would be inflated if evaluation requests leaked into training.

Editorial extensions

If this is right

If the central claim is correct, base stations can run a diffusion-based popularity predictor without requiring GPUs at the user side, since the local model has only 770K parameters.
Cache hit percentage increases with the number of users participating in federated training, so broadening participation is a direct performance lever for the scheme.
Diffusion time steps show diminishing returns: the reported gains saturate near $T=50$, and larger $T$ mostly raises training time, giving a practical tuning rule for deployment.
The scheme's privacy protection rests on exchanging model updates instead of raw ratings, so it inherits the usual federated-learning assumption that model updates do not reveal individual user data.

Reading between the lines

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

Beyond the paper's claims, the same latent-space federated diffusion pipeline could be tested on implicit feedback logs such as browsing or watch histories, since MovieLens 1M consists of explicit ratings that may be easier to predict than clickstream data.
The popularity score averages all generated samples, so the model could in principle be repurposed for personalized per-user ranking; the paper only claims aggregate popularity prediction.
A natural stress test the paper does not run is a privacy attack that tries to recover individual users' rated items from uploaded local models or generated samples, which would sharpen the privacy claim.
The 16-dimensional latent bottleneck suggests the scheme learns a compact embedding of user taste, a property that could transfer to recommendation systems beyond edge caching.

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.

Desk Editor's Note

read the letter

Colleague,

Quick take on arXiv:2506.04593. The paper combines federated learning with a lightweight latent-space diffusion model for edge caching. That is genuinely new in the edge-caching literature: prior work used GANs or deep RL, and this is the first to make a 770K-parameter 1D-conv U-Net LDPM work inside an FL loop for popularity prediction. The diffusion equations are standard and correctly reproduced. The system model is clear, the figures of merit (cache hit percentage and delay) are the right ones for the subfield, and the source code is released. Credit where due: the authors did the work of making the model small enough to matter for resource-constrained devices, and the comparison against CPPPP, Thompson Sampling, and raw LDPM is the right baseline set.

The soft spots are real but mostly fixable. First, the evaluation never states the MovieLens train/test split. The paper says the encoder/decoder are pretrained "on the BS using publicly available datasets," but does not say whether that includes MovieLens, which part, or how the 1M ratings were converted to sparse vectors. Since cache hit percentage is measured on user requests, if the encoder or LDPM saw those requests during training, the reported gains are inflated and the raw-LDPM comparison is unfair. MovieLens has timestamps, so a temporal split is possible; it should have been described. Second, Eq. (9) looks wrong: an FL aggregation step normally sets the next global model to a weighted average of local models, not ω^r minus a weighted sum. Either it's a typo or the algorithm does something unusual; it needs fixing. Third, the privacy claim is too strong. FL reduces raw-data exposure, but model updates can leak user information; the paper asserts privacy without an attack model or defense analysis. That is a minor overclaim for a letter, but worth tightening. Fourth, no error bars or repeated runs; Table II gives single numbers.

None of this kills the central idea. The contribution is incremental but real, and the missing split is an evaluation gap, not a mathematical contradiction. With a clarified experimental protocol and a corrected aggregation equation, this is publishable as a letter.

Who it is for: readers working on edge caching with generative models; it deserves a serious referee. Bring it to a reading group if you work in FL+caching, otherwise a quick skim suffices.

Recommendation: send it to peer review, with the expectation that the authors will need to state the data split and correct Eq. (9).

Referee Report

4 major / 5 minor

Summary. The paper proposes a federated-learning-assisted edge caching scheme based on a lightweight denoising diffusion probabilistic model (LDPM). Users train a small 1D-convolutional U-Net diffusion model locally on latent representations of their request data produced by a pretrained encoder, and the base station aggregates the local models and uses the global model to generate synthetic user data for popularity prediction. The authors claim higher cache hit percentage and lower request delay than CPPPP, Thompson Sampling, and a raw-data LDPM baseline on MovieLens 1M, while preserving privacy through federated training. A source-code repository is provided.

Significance. If the empirical claims are reliable, the contribution is a reasonable and timely combination of a compact diffusion model (770K parameters) with federated learning for edge caching, and the latent-space treatment of sparse user data is a sensible design choice. The paper also addresses a real constraint: edge devices cannot train large generative models. A notable strength is the released source code, which improves reproducibility. However, the comparative advantage over existing FL-based methods rests entirely on the simulation section, and the evaluation as described is too underspecified to support the headline conclusion. The missing train/test split, the unclear pretraining data for the encoder/decoder, the questionable aggregation formula, and the lack of statistical reporting are all load-bearing for the central claim. The conceptual framework is plausible, but the current evidence is not yet convincing.

major comments (4)

[Section IV and Table I] The simulation section never states how the MovieLens 1M dataset is split into training and evaluation requests, nor how each rating is converted into the 3952-dimensional input vector (binarized presence, normalized 1–5 rating, or something else). Table I lists only the encoder structure dimensions. This matters because every component that influences the predicted popularity scores—the pretrained encoder/decoder and the federated LDPM—must be trained only on past requests, with cache hits computed on future requests. The text in Section III says the encoder/decoder are "trained on the BS using publicly available datasets" without specifying which dataset or which portion; given that the encoder input dimension equals the MovieLens item count, the risk that pretraining data overlaps the evaluation requests is concrete. If the encoder/decoder or the LDPM saw the evaluation requests, the popularity scores in Eq. (10) would be fitted to the test data and the gains in Figs. 3–5 would be inflated. The authors should specify the exact split (ideally temporal, using MovieLens timestamps), the rating encoding, and the pretraining dataset, and confirm that none of the evaluation requests were used in any training stage.
[Section III-A, Eq. (9)] The model aggregation formula is not the weighted average described in the text. Eq. (9) reads ω^{r+1} = ω^r − η Σ_{i} (|d_i|/d) ω_i^r, which subtracts a weighted combination of local model parameters from the current global model. In standard FedAvg, the local models are initialized from the global model, so the aggregated global model should be a weighted average of the local models (or the global model plus a weighted average of model deltas). As written, the equation is dimensionally inconsistent with the stated operation and cannot produce the new global model that the subsequent content-prediction step relies on. This is not a purely notational issue because the aggregation step is central to the federated training algorithm. The equation should be corrected to a weighted average, or the notation should be changed if ω_i^r denotes an update or gradient rather than a model.
[Section IV, Figs. 3–5 and Table II] The evaluation reports single curves and single numbers with no error bars, no repeated runs, and no seed information. The scheme involves stochastic components—minibatch sampling in Eq. (7), random noise in the diffusion process, and the U generated samples in Eq. (10)—so point estimates are not sufficient to establish that the proposed scheme is better than CPPPP or Thompson Sampling, or that the differences between T values in Table II are meaningful. The authors should report mean and variance over multiple independent runs, or at minimum provide a statistical test for the key comparisons. This is particularly important in Fig. 4, where the delay differences between schemes appear small.
[Section IV, baseline descriptions] The baseline implementations are not specified in enough detail to judge fairness. For CPPPP, the paper does not state the architecture, learning rate, number of local iterations, or whether it uses the same user set and the same data split as the proposed scheme. For Thompson Sampling, the implementation details (e.g., prior, number of arms, update schedule) are omitted. The raw-LDPM baseline is also an ablation rather than a comparable alternative, because it intentionally does not use the encoder/decoder, and the paper's conclusion that it performs worst is therefore expected. The comparison would be more convincing if the authors reported the tuning effort and hyperparameters for each baseline, and ideally showed that the proposed advantage is robust to reasonable variations in baseline training budgets.

minor comments (5)

[Section II-B, Eqs. (2)–(3)] The symbol ᾱ_t is defined but not used in Eqs. (2) and (3); the forward-process equations should use ᾱ_t (the cumulative product) rather than α_t to match the standard DDPM formulation and the stated definition.
[Section III-B] There is a typographical error: "dncoder" should be "decoder."
[Section IV] The text says each MovieLens rating ranges from 0 to 5, but MovieLens 1M ratings are on a 1–5 scale; this should be corrected, and the actual encoding used should be stated.
[Section IV, Table I] The caption of Table I should list all parameters used in the experiments; hyperparameters such as the number of federated rounds R_max, the diffusion model learning rate schedule, and the batch size are not given.
[General] Some figure labels in the PDF are garbled (e.g., Fig. 3 legend strings beginning with "/uni00000032"), which makes the figures hard to read; the figure source files should be regenerated with standard text encoding.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity in the derivation; the LDPM training and popularity scoring are defined by standard equations, and any concern about the MovieLens split is a data-leakage risk, not a demonstrated circularity.

full rationale

I walked the claimed derivation chain from the diffusion equations (1)-(6) through the FL local loss (7), the local update (8), model aggregation (9), and the content popularity scoring in Eq. (10). Each step is specified by standard formulas rather than by the cache-hit metric being predicted; Eq. (10) explicitly defines the popularity score as the average of generated samples, not as a hidden restatement of the evaluation outcome. The pre-trained encoder/decoder and the 1D U-Net are architectural choices drawn from latent-diffusion and lightweight-diffusion literature, and no uniqueness claim or fitted parameter is later renamed as a prediction. The possibly overlapping-author references [7], [11], and [13] serve as background, motivation, or an experimental comparator (CPPPP); none carries the argument in a load-bearing way, and no prior theorem by the same authors is invoked to force the design. The main empirical concern is that Section IV does not specify the MovieLens temporal split or the dataset used to pre-train the encoder/decoder; if the models saw the evaluation requests, the reported cache-hit gains could be inflated. That is a data-leakage/reproducibility risk, however, not a circularity demonstrated by an equation-level reduction, and the instructions require an exhibited reduction before flagging circularity. I therefore find no significant circularity.

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

The central claim rests on the validity of the DPM objective, the fidelity of the latent-space autoencoder, the Monte Carlo popularity estimate, and the convergence of federated averaging. The first is standard; the last three are domain assumptions with limited direct evidence in the paper. Free parameters are the diffusion time steps, sample count, learning rate, local iterations, and latent dimension, none of which are rigorously justified.

free parameters (5)

U (generated sample count) = 1000
Number of samples drawn from the global LDPM for popularity scoring; chosen without sensitivity analysis, affects variance of the popularity estimate (Section IV, Table I).
e (local training iterations) = 30
Number of local gradient steps per FL round; set as a hyperparameter without tuning discussion (Section IV, Table I).
eta_d (LDPM learning rate) = 0.0006
Local model learning rate; fixed in Table I, no grid search reported.
Latent dimension of encoder = 16
The pretrained encoder compresses 3952-dim rating vectors to 16 dims (Table I); this architectural choice determines the information bottleneck and is not swept.
Diffusion time steps T = 50
Selected after a sweep in Table II: T=50 gives near-asymptotic cache hit with moderate training time; higher T adds cost without gain.

assumptions (4)

standard math The DPM forward/reverse process and simplified loss L_simple from Ho et al. (Eq. 6) is a valid generative model for the user data distribution in latent space.
The paper builds on [4]; no derivation of the DPM objective is given, it is taken as background.
domain assumption The pretrained encoder/decoder, trained on the BS with publicly available datasets, provides a faithful latent representation of the high-dimensional sparse user rating vectors.
Section III states this as the approach (encoder maps d_i to 16-dim latent), but no validation of reconstruction quality or representational sufficiency is given.
domain assumption The average of U decoded generated samples approximates the true popularity ranking of contents in the test period.
Section III-B averages generated samples and caches top-N; the validity of this Monte Carlo estimate as a popularity predictor is assumed and only indirectly tested via cache hit percentage.
domain assumption Federated averaging of local diffusion model parameters (Eq. 9 or an unstated correction) yields a global model that captures the aggregate content preferences of all users.
The paper assumes the FL algorithm converges to a useful global generator; no convergence analysis or comparison against centralized training is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Federated Learning Assisted Edge Caching Scheme Based on Lightweight Architecture DDPM." pith.science (2026). https://pith.science/paper/SRAGFA3J

@misc{pith2026250604593,
  author       = {Pith},
  title        = {Pith review of: Federated Learning Assisted Edge Caching Scheme Based on Lightweight Architecture DDPM},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SRAGFA3J}},
  note         = {Machine review of arXiv:2506.04593}
}

read the original abstract

Edge caching is an emerging technology that empowers caching units at edge nodes, allowing users to fetch contents of interest that have been pre-cached at the edge nodes. The key to pre-caching is to maximize the cache hit percentage for cached content without compromising users' privacy. In this letter, we propose a federated learning (FL) assisted edge caching scheme based on lightweight architecture denoising diffusion probabilistic model (LDPM). Our simulation results verify that our proposed scheme achieves a higher cache hit percentage compared to existing FL-based methods and baseline methods.

Figures

Figures reproduced from arXiv: 2506.04593 by the authors.

**Figure 1.** System Model. B. Denoising Diffusion Probabilistic Model (DPM) The theoretical foundation of diffusion models stems from the entropy increase-inverse process of non-equilibrium thermodynamic systems. DPM achieves forward diffusion process and reverse diffusion process through parameterized Markov chains. 1) Forward Diffusion Process: By parameterizing a Markov chain with a scheduling strategy {βt} T t=1, Gaussian n… view at source ↗

**Figure 2.** Encoder, Decoder and LDPM framework. III. EDGE CACHING SCHEME This section introduces the proposed edge caching scheme. We first introduce the federated-based LDPM training algorithm, and then introduce the content popularity prediction algorithm. Additionally, user data is typically high-dimensional and sparse, causing the Euclidean distances between data points to become uniform, and the noise distribution to bec… view at source ↗

**Figure 4.** Request content delay versus cache capacity. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

Figures from the paper (2 more)

**Figure 3.** Figure 3: Cache hit percentage versus cache capacity. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png]

**Figure 5.** Figure 5: Cache hit percentage and request content delay versus the [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

18 extracted references · 11 canonical work pages

[1]

A Dynamical Framework for Modeling Fear of Infection and Frustration with Social Distancing in COVID-19 Spread

J. Zhang, P. Fan, and K. B. Letaief, “Network Coding for Ef- ficient Multicast Routing in Wireless Ad-hoc Networks,”IEEE Trans. Commun., vol. 56, no. 4, pp. 598–607, Apr. 2008, doi: 10.1109/TCOMM.2008.060238

work page Pith review arXiv 2008
[2]

Pricing and Resource Allocation via Game Theory for a Small-Cell Video Caching System,

J. Li, H. Chen, Y . Chen, Z. Lin, B. Vucetic, and L. Hanzo, “Pricing and Resource Allocation via Game Theory for a Small-Cell Video Caching System,”IEEE J. Select. Areas Commun., vol. 34, no. 8, pp. 2115–2129, Aug. 2016, doi: 10.1109/JSAC.2016.2577278

work page arXiv 2016
[3]

Context- Aware Proactive Content Caching With Service Differentiation in Wireless Networks,

S. Muller, O. Atan, M. V . D. Schaar, and A. Klein, “Context- Aware Proactive Content Caching With Service Differentiation in Wireless Networks,”IEEE Trans. Wireless Commun., vol. 16, no. 2, pp. 1024–1036, Feb. 2017, doi: 10.1109/TWC.2016.2636139

work page arXiv 2017
[4]

Denoising diffusion probabilistic models,

J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” inProc. Adv. Neural Inf. Process. Syst., vol. 33, 2020, pp. 6840–6851

2020
[5]

Snapfusion: Text-to-image diffusion model on mobile devices within two seconds,

Y . Li, H. Wang, Q. Jin, J. Hu, P. Chemerys, Y . Fu, Y . Wang, S. Tulyakov, and J. Ren, “Snapfusion: Text-to-image diffusion model on mobile devices within two seconds,”NeurIPS, 2024

work page 2024
[6]

LightGrad: Lightweight Diffusion Probabilistic Model for Text-to-Speech,

J. Chen, X. Song, Z. Peng, B. Zhang, F. Pan, and Z. Wu, “LightGrad: Lightweight Diffusion Probabilistic Model for Text-to-Speech,” in ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Rhodes Island, Greece: IEEE, Jun. 2023, pp. 1–5. doi: 10.1109/ICASSP49357.2023.10096710

arXiv 2023
[7]

Cooperative Edge Caching Based on Elastic Federated and Multi- Agent Deep Reinforcement Learning in Next-Generation Networks,

Q. Wu, W. Wang, P. Fan, Q. Fan, H. Zhu, and K. B. Letaief, “Cooperative Edge Caching Based on Elastic Federated and Multi- Agent Deep Reinforcement Learning in Next-Generation Networks,” IEEE Trans. Netw. Serv. Manage., vol. 21, no. 4, pp. 4179–4196, Aug. 2024

work page 2024
[8]

Adaptive Federated Learning in Resource Constrained Edge Computing Systems,

S. Wang, T. Tuor, T. Salonidis, K. K. Leung, C. Makaya, T. He and K. chan, “Adaptive Federated Learning in Resource Constrained Edge Computing Systems,”IEEE J. Select. Areas Commun., vol. 37, no. 6, pp. 1205–1221, Jun. 2019, doi: 10.1109/JSAC.2019.2904348

work page arXiv 2019

Show all 18 references

[9]

Federated Deep Reinforcement Learning for Internet of Things With Decentral- ized Cooperative Edge Caching,

X. Wang, C. Wang, X. Li, V . C. M. Leung, and T. Taleb, “Federated Deep Reinforcement Learning for Internet of Things With Decentral- ized Cooperative Edge Caching,”IEEE Internet Things J., vol. 7, no. 10, pp. 9441–9455, Oct. 2020, doi: 10.1109/JIOT.2020.2986803

2020
[10]

Mobility- Aware Proactive Edge Caching for Connected Vehicles Using Federated Learning,

Z. Yu, J. Hu, G. Min, Z. Zhao, W. Miao, and M. S. Hossain, “Mobility- Aware Proactive Edge Caching for Connected Vehicles Using Federated Learning,”IEEE Trans. Intell. Transport. Syst., vol. 22, no. 8, pp. 5341–5351, Aug. 2021, doi: 10.1109/TITS.2020.3017474

2021
[11]

Mobility- Aware Cooperative Caching in Vehicular Edge Computing Based on Asynchronous Federated and Deep Reinforcement Learning,

Q. Wu, Y . Zhao, Q. Fan, P. Fan, J. Wang, and C. Zhang, “Mobility- Aware Cooperative Caching in Vehicular Edge Computing Based on Asynchronous Federated and Deep Reinforcement Learning,”IEEE J. Sel. Top. Signal Process., vol. 17, no. 1, pp. 66–81, Jan. 2023, doi: 10.1109/JSTSP...

2023
[12]

Asyn- chronous Federated and Reinforcement Learning for Mobility-Aware Edge Caching in IoV ,

K. Jiang, Y . Cao, Y . Song, H. Zhou, S. Wan, and X. Zhang, “Asyn- chronous Federated and Reinforcement Learning for Mobility-Aware Edge Caching in IoV ,”IEEE Internet Things J., vol. 11, no. 9, pp. 15334–15347, May 2024, doi: 10.1109/JIOT.2023.3349255

2024
[13]

An Efficient Content Popularity Prediction of Privacy Preserving Based on Federated Learning and Wasserstein GAN,

K. Wang, N. Deng, and X. Li, “An Efficient Content Popularity Prediction of Privacy Preserving Based on Federated Learning and Wasserstein GAN,”IEEE Internet Things J., vol. 10, no. 5, pp. 3786–3798, Mar. 2023, doi: 10.1109/JIOT.2022.3176360

2023
[14]

RadioDiff: An Effective Generative Diffusion Model for Sampling- Free Dynamic Radio Map Construction,

X. Wang, K. Tao, N. Cheng, Z. Yin, Z. Li, Y . Zhang and X. Shen, “RadioDiff: An Effective Generative Diffusion Model for Sampling- Free Dynamic Radio Map Construction,”IEEE Trans. Cogn. Commun. Netw., pp. 1–1, 2025, doi: 10.1109/TCCN.2024.3504489

2025
[15]

High- Resolution Image Synthesis with Latent Diffusion Models,

R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- Resolution Image Synthesis with Latent Diffusion Models,” in2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, USA: IEEE, Jun. 2022, pp. 10674–10685. doi: 10.1109/CVPR...

2022
[16]

U-Net: Convolutional net- works for biomedical image segmentation,

O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional net- works for biomedical image segmentation,” inProc. Int. Conf. Med. Image Comput. Comput.-Assist. Intervent., 2015, pp. 234–241

2015
[17]

Grad-tts: A diffusion probabilistic model for text-to-speech,

Vadim Popov, Ivan V ovk, Vladimir Gogoryan, Tasnima Sadekova, and Mikhail A. Kudinov, “Grad-tts: A diffusion probabilistic model for text-to-speech,” inICML, 2021, vol. 139, pp. 8599–8608

2021
[18]

DiffWave: Aversatile diffusion model for audio synthesis,

Z. Kong, W. Ping, J. Huang, K. Zhao, and B. Catanzaro, “DiffWave: Aversatile diffusion model for audio synthesis,” inProc. Int. Conf. Learn. Representations, 2021

2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

A Dynamical Framework for Modeling Fear of Infection and Frustration with Social Distancing in COVID-19 Spread

J. Zhang, P. Fan, and K. B. Letaief, “Network Coding for Ef- ficient Multicast Routing in Wireless Ad-hoc Networks,”IEEE Trans. Commun., vol. 56, no. 4, pp. 598–607, Apr. 2008, doi: 10.1109/TCOMM.2008.060238

work page Pith review arXiv 2008

[2] [2]

Pricing and Resource Allocation via Game Theory for a Small-Cell Video Caching System,

J. Li, H. Chen, Y . Chen, Z. Lin, B. Vucetic, and L. Hanzo, “Pricing and Resource Allocation via Game Theory for a Small-Cell Video Caching System,”IEEE J. Select. Areas Commun., vol. 34, no. 8, pp. 2115–2129, Aug. 2016, doi: 10.1109/JSAC.2016.2577278

work page arXiv 2016

[3] [3]

Context- Aware Proactive Content Caching With Service Differentiation in Wireless Networks,

S. Muller, O. Atan, M. V . D. Schaar, and A. Klein, “Context- Aware Proactive Content Caching With Service Differentiation in Wireless Networks,”IEEE Trans. Wireless Commun., vol. 16, no. 2, pp. 1024–1036, Feb. 2017, doi: 10.1109/TWC.2016.2636139

work page arXiv 2017

[4] [4]

Denoising diffusion probabilistic models,

J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” inProc. Adv. Neural Inf. Process. Syst., vol. 33, 2020, pp. 6840–6851

2020

[5] [5]

Snapfusion: Text-to-image diffusion model on mobile devices within two seconds,

Y . Li, H. Wang, Q. Jin, J. Hu, P. Chemerys, Y . Fu, Y . Wang, S. Tulyakov, and J. Ren, “Snapfusion: Text-to-image diffusion model on mobile devices within two seconds,”NeurIPS, 2024

work page 2024

[6] [6]

LightGrad: Lightweight Diffusion Probabilistic Model for Text-to-Speech,

J. Chen, X. Song, Z. Peng, B. Zhang, F. Pan, and Z. Wu, “LightGrad: Lightweight Diffusion Probabilistic Model for Text-to-Speech,” in ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Rhodes Island, Greece: IEEE, Jun. 2023, pp. 1–5. doi: 10.1109/ICASSP49357.2023.10096710

arXiv 2023

[7] [7]

Cooperative Edge Caching Based on Elastic Federated and Multi- Agent Deep Reinforcement Learning in Next-Generation Networks,

Q. Wu, W. Wang, P. Fan, Q. Fan, H. Zhu, and K. B. Letaief, “Cooperative Edge Caching Based on Elastic Federated and Multi- Agent Deep Reinforcement Learning in Next-Generation Networks,” IEEE Trans. Netw. Serv. Manage., vol. 21, no. 4, pp. 4179–4196, Aug. 2024

work page 2024

[8] [8]

Adaptive Federated Learning in Resource Constrained Edge Computing Systems,

S. Wang, T. Tuor, T. Salonidis, K. K. Leung, C. Makaya, T. He and K. chan, “Adaptive Federated Learning in Resource Constrained Edge Computing Systems,”IEEE J. Select. Areas Commun., vol. 37, no. 6, pp. 1205–1221, Jun. 2019, doi: 10.1109/JSAC.2019.2904348

work page arXiv 2019