Pith. sign in

REVIEW 5 major objections 5 minor 27 references

Secure Cluster-Based Hierarchical Federated Learning in Vehicular Networks

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A reliability-scoring defense keeps vehicular federated learning within 2–3% of ideal accuracy under poisoning attacks.

desk verdict Plausible integration of known defenses with a solid ablation, but the headline numbers don't reproduce and the reliability scoring rests on an unflagged labeled validation set at the server. read the letter →

arxiv 2505.01186 v1 pith:IAGKQ6YF submitted 2025-05-02 cs.CR cs.AIcs.DCcs.LGcs.SYeess.SY

M. Saeid HaghighiFard , Sinem Coleri This is my paper · ORCID
classification cs.CRcs.AIcs.DCcs.LGcs.SYeess.SY
keywords hierarchicalfederatedlearningvehicularnetworksmodelpoisoningdefenseanomalydetectiondynamicclientselectionGaussiannoiseattackgradientascentadaptivethresholding
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 proposes a defense, DARCS, for hierarchical federated learning in vehicular networks, and claims it keeps the global model within 2–3% of attack-free accuracy while cutting convergence time by up to 17% compared with Z-score-only, cosine-similarity-only, and combined defenses. The defense targets two poisoning styles: Gaussian noise added to local updates, and gradient-ascent updates that push the model away from the loss minimum. It scores every vehicle on historical accuracy, contribution frequency, and anomaly history, then uses Z-score and cosine-similarity checks, adaptive thresholds, weighted averaging, and cross-cluster consistency checks to filter suspicious contributions. The result matters because vehicular federated learning is attractive for privacy and bandwidth, but an undefended hierarchy gives attackers many places to inject bad updates. If the claim holds, this is a practical recipe for keeping collaborative vehicle models reliable under persistent attack.

What carries the argument

The device carrying the argument is the reliability score, a weighted combination of historical accuracy, contribution frequency, and anomaly record (Eq. 15), used both to select participating vehicles and to weight their gradients in aggregation (Eqs. 17 and 19). Two statistical detectors feed it: a Z-score on gradient norms (Eq. 4) to catch magnitude outliers typical of Gaussian noise, and a cosine similarity against the mean update (Eqs. 6 and 7) to catch directional reversals typical of gradient ascent. An adaptive threshold (Eq. 16) tightens the cosine criterion for high-accuracy vehicles, and a cross-cluster consistency check (Eqs. 8 and 9) compares each cluster's update with all others to expose coordinated attacks. These mechanisms jointly determine which updates enter the weighted average, so the defense's power rests on the accuracy of the reliability scores.

What would settle it

Withhold the labeled validation set from cluster heads and the central server, keep 20 percent of vehicles attacking from round one, and measure accuracy and convergence time; if the model stays within 3 percent of attack-free accuracy, the validation-set assumption is not load-bearing, and if it does not, the defense's central premise is exposed.

Watch

Extended reading notes

Core claim

The paper's central claim is that a multi-level reliability-based filter can defend hierarchical federated learning in vehicular networks against both noise injection and gradient-ascent model poisoning without sacrificing convergence speed. At each round, cluster heads and the central EPC compute reliability scores for each vehicle from its historical accuracy on a validation dataset, its contribution frequency, and its anomaly record; Z-score analysis catches updates whose norms deviate statistically, while cosine similarity catches updates whose direction disagrees with consensus. An adaptive threshold tightens cosine-similarity scrutiny for vehicles with high historical accuracy, and a cross-cluster consistency check flags clusters whose aggregated updates diverge from the rest of the network. Accepted updates are averaged with weights proportional to reliability, so trusted vehicles dominate aggregation. Simulations across 1-hop and 3-hop topologies, with 20% of vehicles attacking from round one, show DARCS holds accuracy loss to roughly 2–3% of the attack-free baseline while benchmarks lose 8–17% and sometimes fail to converge at tight thresholds.

Load-bearing premise

The defense assumes the central server and every cluster head hold a labeled validation dataset on which to score each vehicle's model, yet standard federated learning gives the server no such labeled data.

Editorial extensions

If this is right

  • Vehicular HFL deployments could run under sustained Gaussian-noise and gradient-ascent attacks with convergence times near the attack-free baseline rather than diverging.
  • The 3-hop topology inherits an extra aggregation layer that dilutes random noise, so DARCS converges slightly faster there than in 1-hop under the same attacks.
  • Adaptive thresholding alone recovers 4–6% of accuracy compared with static thresholds, and cross-cluster consistency checks recover a further 3–5%.
  • Benchmark defenses that use only Z-score, only cosine similarity, or a fixed combination will fail to converge at tight thresholds, while DARCS continues to converge.

Reading between the lines

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

  • A testable extension would drop the labeled validation dataset from the cluster heads and EPC, since ordinary federated learning has no such server-side labels; the accuracy gap would likely widen and expose how much of the defense depends on that supervision.
  • The same reliability-scoring structure could be adapted to other poisoning styles, such as backdoor or label-flipping attacks, but cosine similarity against consensus would be weaker against attacks designed to stay close to the benign direction.
  • The adaptive threshold that tightens scrutiny for high-performing vehicles could over-exclude honest vehicles after a natural data-distribution shift, since their historical accuracy would briefly drop and their updates would look anomalous.
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

5 major / 5 minor

Summary. The paper proposes DARCS, a defense framework for cluster-based hierarchical federated learning (HFL) in vehicular networks. DARCS combines Z-score and cosine-similarity anomaly detection with per-vehicle reliability scores built from historical accuracy, contribution frequency, and anomaly records; it adds an adaptive thresholding mechanism for cosine similarity, a cross-cluster consistency check, and reliability-weighted gradient averaging. The authors evaluate DARCS on an MNIST-based non-IID vehicular simulation with 20% attackers under Gaussian noise, gradient ascent, and combined attacks, in both 1-hop and 3-hop topologies, comparing against cosine-similarity-only, Z-score-only, and a combined baseline. The paper claims that DARCS limits accuracy loss to within 2–3% of the attack-free baseline and reduces convergence time by up to 17% relative to the benchmarks.

Significance. If the claims hold, DARCS would be a practically useful, multi-level defense for hierarchical federated learning in vehicular settings, and the paper addresses a genuine gap: existing defenses are not designed for the cluster-based HFL structure. The algorithm descriptions are unusually complete, with pseudocode for both the CH and EPC levels, and the evaluation spans multiple attack types, network sizes, transmission ranges, and convergence thresholds. However, the central quantitative claims are not reproducible from the reported tables, and a key input—labeled validation data at the aggregators—is never stated as an assumption. The value of the contribution therefore depends on whether these issues can be resolved with additional experiments and a clearer statement of the threat and system model.

major comments (5)
  1. [Section III.A.3, Eqs. (10)–(11), Algorithms 1 and 2] Historical Accuracy and every Accuracy of Contribution computation require a labeled validation set D_val at the CH and EPC. The paper never lists this as an assumption. In standard federated learning, the aggregator does not hold labeled client data; without D_val, Eq. (10) is undefined, Eq. (15) and the adaptive-thresholding trigger cannot be evaluated, and Eqs. (17) and (19) reduce to unweighted averaging over the Z-score/cosine-filtered updates. The claimed 2–3% accuracy loss and up to 17% faster convergence are therefore demonstrated only under an unstated availability of validation data. Please state this assumption explicitly, justify its availability in the vehicular HFL setting, or provide experiments without D_val. Even when D_val exists, using per-client accuracy on a fixed D_val as a trust signal can systematically favor clients whose local data distribution resembles D_val under non-IID data, so the validity of the metric needs discussion.
  2. [Abstract and Section V vs. Tables I and II] The abstract states that under combined noise and gradient ascent in the 1-hop scenario DARCS reduces convergence time by 17.1% compared to cosine-similarity and Z-score defenses and by 8.7% over the combined defense. These percentages do not appear in the combined-attack rows of Table I; they match the noise-only rows instead (DARCS 63 vs. Z-score 76 = 17.1%, and DARCS 63 vs. combined 69 = 8.7%, at ε=0.01). For the combined-attack row at ε=0.01, the improvements are 12.2% over cosine similarity (65 vs. 74) and 7.1% over the combined defense (65 vs. 70). The 3-hop percentages (15.6% and 7.1%) also do not match a single row of Table II. Please recompute the headline numbers and cite the exact table rows, or revise the claims.
  3. [Section III.A.4, Eq. (16), Algorithm 1 lines 26–27, Algorithm 2 lines 32–33] Equation (16) defines the adaptive threshold as itself minus δ. As written, this is not a mathematical definition; if it is intended as an iterative update rule, the initial value and the update schedule are missing. The algorithms do not specify how many times δ is applied per round, whether the threshold is per-vehicle state, or what happens at the floor High Threshold Down. Since adaptive thresholding is credited with 4–6% accuracy improvement in Section V, this mechanism needs a well-defined recurrence with an explicit initial condition and update trigger.
  4. [Section IV, Tables I and II] The performance comparison is based on single runs. The mobility model, data partitioning, and client selection contain randomness, but no standard deviations, confidence intervals, or numbers of seeds are reported. Many of the claimed improvements are 2–7 rounds out of roughly 60–80 rounds, which may be within run-to-run variation. Please report repeated-seed experiments with error bars and a sensitivity analysis over Z Score Threshold, δ, High Threshold Up/Down, and the reliability-score weights. This is needed to support the statements that DARCS “consistently outperforms” the benchmarks and achieves performance “remarkably close” to the no-attack baseline.
  5. [Eqs. (15), (17), and (19)] The reliability score in Eq. (15) is a weighted sum with a negative anomaly term but no non-negativity constraint. The weighted averaging in Eqs. (17) and (19) divides by the sum of reliability scores; if any score is negative or the total sum is non-positive, the aggregation is ill-defined or the update direction can be inverted. Please specify the allowed range of reliability scores and the normalization strategy, or prove that the scores remain positive under the proposed update rules.
minor comments (5)
  1. [Section I, last paragraph] The introduction says that Section VI presents concluding remarks, but the conclusion is actually Section V.
  2. [Algorithm 1, lines 8–9] The sorting and selection of Selected Clients appear inside the per-CM foreach loop due to the indentation; these operations should be performed once per round, not once per CM.
  3. [Algorithm 2, line 10 and surrounding text] The text says the CH cosine similarity compares each CH’s update with the global model parameters, but Eq. (7) defines a temporal consistency between the current and previous aggregated updates relative to the previous global model; these descriptions should be reconciled.
  4. [Section IV.B vs. Section IV.C] The accuracy-loss figures are inconsistent: Section IV.B says adaptive thresholding limits loss to within 1–3% of baseline, while Section IV.C and the conclusion say 2–3%; please use one consistent range.
  5. [Section IV.A] The reliability-score weights are described only as “equal”; please provide the exact weight values and the accuracy/frequency/anomaly normalization in the setup table.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the DARCS claims are supported by simulation against external benchmark defenses, and the validation-set oracle is an assumption, not a construction-level circularity.

full rationale

The central claims (up to 17% faster convergence, 2-3% accuracy loss) are empirical results from the simulations in Section IV, compared against external benchmark defenses (CosDefense [15], Z-Score [13], and a combined variant). The paper does not derive these numbers from its own assumptions by an equation-level reduction. The closest candidate is the validation-set dependence: Eq. (10)-(11) define Historical Accuracy from Accuracy of Contribution on D_val, Eq. (15) builds Reliability Score from it, and Eqs. (17)/(19) use that score for weighted averaging. This does mean the defense assumes a labeled validation dataset at the CH and EPC, which is a significant correctness/privacy assumption in federated learning, and if the final evaluation accuracy were measured on the same D_val the reported accuracy would be partly self-fulfilling. However, the paper does not state that the final evaluation metric equals accuracy on D_val; evaluation accuracy on a separate held-out set is the standard reading, and no quoted equation forces the reported outcomes from the validation-set inputs. The adaptive threshold update in Eq. (16) is self-referential in form, but Section IV.A supplies an initial value (Cosine Sim Threshold adaptive,k = 0.90), making it an update rule rather than a definition that assumes its own conclusion. The self-citation [6] provides the CbHFL clustering substrate, but it is not the load-bearing element of the defense claim; the benchmarks and ablations isolate the DARCS mechanisms. No uniqueness theorem, ansatz, or renamed known result is used to force the conclusion. The validation-data oracle and hand-tuned thresholds are legitimate correctness risks, but they do not constitute circularity under the stated criteria.

Assumptions & free parameters 9 free parameters · 6 assumptions · 0 invented entities

The central claim depends on several hand-set parameters and the availability of labeled validation data at the server, neither of which is derived or justified. The honest-majority assumption underpins both the anomaly detection means and the cross-cluster consistency check.

free parameters (9)
  • Z Score Threshold = 3
    Standard 3-sigma rule, but no sensitivity study; directly gates which updates are flagged.
  • Initial Cosine Sim Threshold adaptive = 0.90
    Hand-set starting threshold for cosine anomaly detection; no justification given.
  • High Threshold Up = 95%
    Accuracy level triggering threshold tightening; tied to the 95% baseline accuracy, so it is tuned to the scenario.
  • High Threshold Down = 0.2
    Floor for adaptive threshold; arbitrary.
  • Adjustment Factor delta = 0.05
    Step size for threshold reduction; arbitrary.
  • Cosine Sim Threshold cross = 0.9
    Threshold for cross-cluster consistency; arbitrary.
  • SELECTED CLIENT PERCENTAGE = 75%
    Fraction of vehicles selected per round; no sensitivity analysis.
  • Unblock Time = 5 rounds
    Block duration for flagged vehicles; arbitrary.
  • Reliability Score Weights = equal
    The three metrics (accuracy, frequency, anomaly record) are weighted equally; no justification for equal weighting.
assumptions (6)
  • domain assumption EPC and CHs have access to a labeled validation dataset D_val to compute per-vehicle historical accuracy (Eq. 10-11).
    Standard FL servers usually do not hold labeled data; this assumption is load-bearing for reliability scoring and adaptive thresholding.
  • domain assumption Honest clients and clusters form a majority at each aggregation round.
    Cross-cluster consistency check and mean-based Z-score/cosine references break if attackers are a majority; simulations use 20% attackers.
  • domain assumption The clustering framework from [6] produces stable clusters and cluster heads.
    DARCS inherits its topology from the authors' earlier CbHFL paper; clustering stability is assumed throughout.
  • domain assumption Gaussian noise and gradient ascent are the only attack types considered.
    The defense is designed specifically for these two attacks; other poisons such as label flipping or backdoors are not tested.
  • domain assumption Z-score and cosine similarity are computed against a mean that approximately represents the benign population.
    Adversarial updates are assumed not to dominate the mean, which is true only if attackers are a minority.
  • ad hoc to paper Equation (16) is interpreted as an iterative update rule rather than a mathematical definition.
    As written, Cosine Sim Threshold = Cosine Sim Threshold - delta is circular; the algorithm makes it a repeated decrement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Secure Cluster-Based Hierarchical Federated Learning in Vehicular Networks." pith.science (2026). https://pith.science/paper/IAGKQ6YF

@misc{pith2026250501186,
  author       = {Pith},
  title        = {Pith review of: Secure Cluster-Based Hierarchical Federated Learning in Vehicular Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IAGKQ6YF}},
  note         = {Machine review of arXiv:2505.01186}
}
read the original abstract

Hierarchical Federated Learning (HFL) has recently emerged as a promising solution for intelligent decision-making in vehicular networks, helping to address challenges such as limited communication resources, high vehicle mobility, and data heterogeneity. However, HFL remains vulnerable to adversarial and unreliable vehicles, whose misleading updates can significantly compromise the integrity and convergence of the global model. To address these challenges, we propose a novel defense framework that integrates dynamic vehicle selection with robust anomaly detection within a cluster-based HFL architecture, specifically designed to counter Gaussian noise and gradient ascent attacks. The framework performs a comprehensive reliability assessment for each vehicle by evaluating historical accuracy, contribution frequency, and anomaly records. Anomaly detection combines Z-score and cosine similarity analyses on model updates to identify both statistical outliers and directional deviations in model updates. To further refine detection, an adaptive thresholding mechanism is incorporated into the cosine similarity metric, dynamically adjusting the threshold based on the historical accuracy of each vehicle to enforce stricter standards for consistently high-performing vehicles. In addition, a weighted gradient averaging mechanism is implemented, which assigns higher weights to gradient updates from more trustworthy vehicles. To defend against coordinated attacks, a cross-cluster consistency check is applied to identify collaborative attacks in which multiple compromised clusters coordinate misleading updates. Together, these mechanisms form a multi-level defense strategy to filter out malicious contributions effectively. Simulation results show that the proposed algorithm significantly reduces convergence time compared to benchmark methods across both 1-hop and 3-hop topologies.

Figures

Figures reproduced from arXiv: 2505.01186 by the authors.

Figure 1
Figure 1. Comparison of accuracy at the EPC with and [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Comparison of accuracy at the EPC with and without [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 20 canonical work pages

  1. [1]

    C. Feng, D. Feng, Z. Zhao, G. Min, and H. Duan, Federated Learning with non-IID data in Mobile Edge Computing Systems , ch. 2, pp. 23–48. John Wiley & Sons, Ltd, 2023

    work page 2023

  2. [2]

    Client-edge-cloud hierarchical federated learning,

    L. Liu, J. Zhang, S. Song, and K. B. Letaief, “Client-edge-cloud hierarchical federated learning,” in ICC 2020 - 2020 IEEE International Conference on Communications (ICC) , pp. 1–6, 2020

    2020

  3. [3]

    Hierarchical federated learning across heterogeneous cellular networks,

    M. S. H. Abad, E. Ozfatura, D. GUndUz, and O. Ercetin, “Hierarchical federated learning across heterogeneous cellular networks,” in ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pp. 8866–8870, 2020

    work page 2020

  4. [4]

    Mobility Accelerates Learning: Convergence Analysis on Hierarchical Federated Learning in Vehicular Networks

    T. Chen, J. Yan, Y . Sun, S. Zhou, D. G ¨und¨uz, and Z. Niu, “Mobility accelerates learning: Convergence analysis on hierarchical federated learning in vehicular networks,” arXiv preprint arXiv:2401.09656, 2024

    work page Pith review arXiv 2024

  5. [5]

    Semi-asynchronous Hierarchical Federated Learning for Cooperative Intelligent Transportation Systems

    Q. Chen, Z. You, and H. Jiang, “Semi-asynchronous hierarchical feder- ated learning for cooperative intelligent transportation systems,” arXiv preprint arXiv:2110.09073, 2021

    work page Pith review arXiv 2021

  6. [6]

    Hierarchical federated learning in multi-hop cluster-based vanets,

    M. S. HaghighiFard and S. Coleri, “Hierarchical federated learning in multi-hop cluster-based vanets,” arXiv preprint arXiv:2401.10361, 2024

    arXiv 2024

  7. [7]

    Poisoning attacks in federated learning: A survey,

    G. Xia, J. Chen, C. Yu, and J. Ma, “Poisoning attacks in federated learning: A survey,” IEEE Access, vol. 11, pp. 10708–10722, 2023

    work page 2023

  8. [8]

    The impact of adversarial attacks on federated learning: A survey,

    K. N. Kumar, C. K. Mohan, and L. R. Cenkeramaddi, “The impact of adversarial attacks on federated learning: A survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence , pp. 1–20, 2023

    work page 2023

Show all 27 references
  1. [9]

    Mpaf: Model poisoning attacks to federated learning based on fake clients,

    X. Cao and N. Gong, “Mpaf: Model poisoning attacks to federated learning based on fake clients,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) , (Los Alamitos, CA, USA), pp. 3395–3403, IEEE Computer Society, jun 2022

    2022

  2. [10]

    Desmp: Differential privacy-exploited stealthy model poisoning attacks in federated learn- ing,

    M. T. Hossain, S. Islam, S. Badsha, and H. Shen, “Desmp: Differential privacy-exploited stealthy model poisoning attacks in federated learn- ing,” in 2021 17th International Conference on Mobility, Sensing and Networking (MSN), pp. 167–174, 2021

    2021

  3. [11]

    Fl- wbc: Enhancing robustness against model poisoning attacks in federated learning from a client perspective,

    J. Sun, A. Li, L. DiValentin, A. Hassanzadeh, Y . Chen, and H. Li, “Fl- wbc: Enhancing robustness against model poisoning attacks in federated learning from a client perspective,” arXiv preprint arXiv:2110.13864 , 2021

    2021 arXiv

  4. [12]

    Poisoning attack detection using client historical similarity in non-iid environments,

    X. You, Z. Liu, X. Yang, and X. Ding, “Poisoning attack detection using client historical similarity in non-iid environments,” in 2022 12th International Conference on Cloud Computing, Data Science & Engineering (Confluence), pp. 439–447, 2022

    2022

  5. [13]

    Defending against data poisoning attacks in federated learning via user elimination,

    N. Galanis, “Defending against data poisoning attacks in federated learning via user elimination,” arXiv preprint arXiv:2404.12778 , 2024

    2024 arXiv

  6. [14]

    Securing federated learning model aggregation against poisoning attacks via credit-based client selection,

    M. Khorramfar, “Securing federated learning model aggregation against poisoning attacks via credit-based client selection,” WinnSpace Home, Aug 2023

    2023

  7. [15]

    Secure federated learning against model poisoning attacks via client filtering,

    D. N. Yaldiz, T. Zhang, and S. Avestimehr, “Secure federated learning against model poisoning attacks via client filtering,” arXiv preprint arXiv:2304.00160, 2023

    2023 arXiv

  8. [16]

    Dedicated short-range communications (dsrc) standards in the united states,

    J. B. Kenney, “Dedicated short-range communications (dsrc) standards in the united states,” Proceedings of the IEEE, vol. 99, no. 7, pp. 1162– 1182, 2011

    2011

  9. [17]

    6g for vehicle-to-everything (v2x) communications: Enabling technologies, challenges, and opportu- nities,

    M. Noor-A-Rahim, Z. Liu, H. Lee, M. O. Khyam, J. He, D. Pesch, K. Moessner, W. Saad, and H. V . Poor, “6g for vehicle-to-everything (v2x) communications: Enabling technologies, challenges, and opportu- nities,” Proceedings of the IEEE , vol. 110, no. 6, pp. 712–734, 2022

    2022

  10. [18]

    Lte evolution for vehicle-to-everything services,

    H. Seo, K.-D. Lee, S. Yasukawa, Y . Peng, and P. Sartori, “Lte evolution for vehicle-to-everything services,” IEEE Communications Magazine , vol. 54, no. 6, pp. 22–28, 2016

    2016

  11. [19]

    5g nr-v2x: Toward connected and cooperative autonomous driving,

    H. Bagheri, M. Noor-A-Rahim, Z. Liu, H. Lee, D. Pesch, K. Moessner, and P. Xiao, “5g nr-v2x: Toward connected and cooperative autonomous driving,” IEEE Communications Standards Magazine , vol. 5, no. 1, pp. 48–54, 2021

    2021

  12. [20]

    Privacy-preserving deep learning,

    R. Shokri and V . Shmatikov, “Privacy-preserving deep learning,” in Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security, pp. 1310–1321, ACM, 2015

    2015

  13. [21]

    Federated learning: Challenges, methods, and future directions,

    T. Li, A. K. Sahu, A. Talwalkar, and V . Smith, “Federated learning: Challenges, methods, and future directions,” IEEE Signal Processing Magazine, vol. 37, no. 3, pp. 50–60, 2019

    2019

  14. [22]

    Analyzing federated learning through an adversarial lens,

    A. N. Bhagoji, S. Chakraborty, P. Mittal, and J. Calo, “Analyzing federated learning through an adversarial lens,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 33, pp. 1360–1367, AAAI, 2019

    2019

  15. [23]

    Anomaly detection: A survey,

    V . Chandola, A. Banerjee, and V . Kumar, “Anomaly detection: A survey,” ACM Computing Surveys (CSUR) , vol. 41, no. 3, pp. 1–58, 2009

    2009

  16. [24]

    Goodfellow, Y

    I. Goodfellow, Y . Bengio, and A. Courville, Deep Learning. MIT Press, 2016

    2016

  17. [25]

    Analytical models of the performance of ieee 802.11p vehicle to vehicle communications,

    M. Sepulcre, M. Gonzalez-Mart ´ın, J. Gozalvez, R. Molina-Masegosa, and B. Coll-Perales, “Analytical models of the performance of ieee 802.11p vehicle to vehicle communications,” IEEE Transactions on Vehicular Technology, vol. 71, no. 1, pp. 713–724, 2022

    2022

  18. [26]

    Study on channel model for frequencies from 0.5 to 100 ghz,

    3rd Generation Partnership Project (3GPP), “Study on channel model for frequencies from 0.5 to 100 ghz,” Tech. Rep. TR 38.901, 3GPP, 2017

    2017

  19. [27]

    Communication-efficient learning of deep networks from decentralized data,

    H. B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, “Communication-efficient learning of deep networks from decentralized data,” in Proceedings of the 20th International Conference on Artificial Intelligence and Statistics , pp. 1273–1282, PMLR, 2017

    2017

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.