Pith. sign in

REVIEW 4 major objections 6 minor 24 references

Fuse and Federate: Enhancing EV Charging Station Security with Multimodal Fusion and Federated Learning

T0 review · 4 major / 6 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 A federated, multimodal intrusion detection framework fusing network traffic and kernel event logs detects attacks on EV charging stations with over 98% accuracy while keeping station data local.

desk verdict The 98%+ detection claim is unsubstantiated: on the CICEVSE2024 network distribution an all-attack classifier already hits ~98.5% accuracy, and the paper reports no majority-class or external baseline. read the letter →

arxiv 2506.06730 v1 pith:UJSBZQFY submitted 2025-06-07 cs.CR cs.AI

Rabah Rahal , Abdelaziz Amara Korba , Yacine Ghamri-Doudane This is my paper
classification cs.CRcs.AI
keywords ElectricVehicleSupplyEquipmentIntrusionDetectionSystemFederatedLearningMultimodalDataFusionNetworkTrafficAnalysisKernelEventsCICEVSE2024EVSEsecurity
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to show that combining two complementary views of an electric vehicle charging station's activity, network traffic logs and low-level kernel and hardware performance events, yields an intrusion detector that is more accurate than either view alone. It proposes a pipeline in which each log type is compressed by its own autoencoder into a compact latent vector, the two vectors are concatenated into a fused representation, and a lightweight 1D CNN classifies the fused vector as benign or malicious. The CNN is trained by federated learning, so stations exchange model updates instead of raw logs. On the CICEVSE2024 dataset the paper reports that the fused local model reaches 92.91% accuracy, and the federated model stays above 98% accuracy and 97% precision across ten simulated stations. The intended payoff is a scalable, privacy-preserving defense for charging infrastructure, which has become an attractive target for ransomware and protocol-level attacks.

What carries the argument

The central object is the fused latent vector $z_i = f_{\text{concat}}(z_i^{(1)}, z_i^{(2)})$, formed by concatenating the bottleneck representations of two modality-specific autoencoders. Each autoencoder $g_j$ compresses a feature vector $x_i^{(j)}$ extracted from one log source into a compact latent code, so fusion combines semantically compressed views rather than raw samples. A one-dimensional CNN maps the fused vector to an intrusion probability, and federated averaging aggregates only the CNN parameter updates across stations. This design is what lets the framework claim multimodal insight with a single lightweight classifier rather than parallel models per modality.

What would settle it

Inspect the data preprocessing to see how network and kernel samples are paired; if the pairing is random or one-to-many, re-run the experiments on timestamp-aligned samples. If accuracy drops to single-modality levels, the reported fusion gain is an artifact of misaligned data rather than genuine multimodal correlation.

Watch

Extended reading notes

Core claim

The paper claims that intermediate fusion of autoencoder latent codes from network traffic and kernel/HPC events improves intrusion detection for EVSE compared with single-modality models, and that federated training preserves most of the centralized accuracy while keeping station data local. On CICEVSE2024, a locally trained fused model reaches 92.91% accuracy versus 92.21% for network-only and 90.54% for kernel-only models. With federated training across ten charging stations, accuracy and F1 remain around 98.9% with a false positive rate of 0.83%, close to the centralized model's near-99% performance. The authors attribute the small gap to local variations in data distribution and present the federated result as a practical balance between privacy and detection quality.

Load-bearing premise

The load-bearing premise is that the 2,000 benign network samples and the 32,303 benign kernel samples can be paired as simultaneous observations of the same station state before their latent vectors are concatenated, but the paper never describes how this matching is done.

Editorial extensions

If this is right

  • EVSE operators could train a shared intrusion detector across many charging stations without shipping raw traffic or kernel logs to a central server.
  • Attacks invisible in network traffic alone, such as local privilege escalation or backdoor installation, become detectable because kernel and HPC events are fused into the same model.
  • The single lightweight CNN and compact latent codes make the approach deployable on resource-constrained edge hardware that cannot run multi-model ensemble detectors.
  • The small accuracy drop between centralized and federated training is presented as the price of privacy, with the federated model still above 98% accuracy across 3 to 10 clients.
  • Because the fusion is modality-agnostic, the same framework extends to power-consumption and voltage logs if those are collected at stations.

Reading between the lines

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

  • The paper does not isolate fusion from federated collaboration: the locally trained fused model scores 92.91% while federated training reaches 98.9%, so part of the headline gain may come from cross-station gradient sharing rather than from multimodal fusion.
  • The privacy claim assumes model updates leak nothing about local data; unless secure aggregation is added, gradient-inversion attacks could recover information, so the practical privacy guarantee is stronger than demonstrated.
  • Algorithm 1's caption mentions on-chain aggregation, but no blockchain or smart-contract component appears in the experiments; the evaluated system is standard federated averaging.
  • Applying the same pipeline to the power-consumption modality already present in CICEVSE2024 would be a direct test of the claimed extensibility, since the design treats modalities as interchangeable autoencoder inputs.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes a federated, multimodal intrusion detection system for electric vehicle supply equipment (EVSE). The framework extracts features from network traffic logs and kernel/HPC event logs, compresses each modality with a dedicated autoencoder, concatenates the latent representations (intermediate fusion), and feeds the fused vector to a 1D CNN trained in a federated manner. Experiments on the CICEVSE2024 dataset report high accuracy (about 98.9% in the federated setting) and claim that the framework outperforms existing solutions. The paper includes comparisons between single-modality and fused models, and between centralized and federated training.

Significance. If the reported results were properly established, the paper would make a useful contribution: it addresses a relevant security problem for EV charging infrastructure, proposes a lightweight multimodal architecture suitable for edge deployment, and leverages federated learning to preserve data privacy. The use of a recent public dataset (CICEVSE2024) and the inclusion of both network- and host-level signals are strengths. However, the central empirical claim—that the framework achieves a detection rate above 98% and outperforms existing solutions—is not supported by the evidence as presented, because no trivial baseline or external comparison is reported and the class-imbalance issue is not addressed. The paper also does not explain how the two modalities are aligned during fusion.

major comments (4)
  1. [§IV-B, Table I, Fig. 3] The headline accuracy and F1-score (about 98.9%) are not compared against a trivial majority-class baseline. Table I shows that the network dataset contains 2,000 benign samples versus 131,580 attack samples (65,790 DoS and 65,790 Recon), so a constant all-attack classifier attains 98.50% accuracy on that distribution. The kernel/HPC dataset has a different benign count (32,303), so the fused distribution is unclear, but the paper never reports the class distribution of the training/test split used for the fused model, nor does it provide a confusion matrix or per-class precision/recall. Without this information, the reported >98% detection rate cannot be interpreted as evidence that the model learned actual attack signatures. The authors should report the test-set class balance, a majority-class baseline, and per-class metrics for each experimental setting.
  2. [§III-B, Eq. (3), Table I] The fusion step concatenates latent vectors z_i^(1) and z_i^(2) for each station i, but Table I shows that the network and kernel/HPC datasets have different numbers of benign samples (2,000 versus 32,303). The paper never explains how samples from the two modalities are aligned or paired into a single fused observation. If the pairing is not based on a shared timestamp or event identifier, the fused vector may combine unrelated network and kernel states, making the reported detection accuracy meaningless. The authors need to specify the alignment procedure, or otherwise justify why unpaired samples can be concatenated.
  3. [§IV-B, Figure 2] The claim that the proposed framework 'outperforms existing solutions' is not supported by the experiments. The only comparisons are against the framework's own single-modality variants (Figure 2) and between centralized and federated versions of the same model (Table III). No external IDS baseline (e.g., a standard machine-learning classifier or a previously published EVSE intrusion detection method) is evaluated on the same data partition. The authors should either add such baselines or temper the claim to say that the framework outperforms its own single-modality variants.
  4. [§III-C, Eq. (7), Algorithm 1] The federated update rule is written as θ_{t+1} = θ_t − η Σ_{i=1}^N ∇θ L(...), which sums client gradients rather than averaging them and does not account for multiple local epochs. As written, the update magnitude grows with the number of clients, which is not the FedAvg algorithm described in the text and would be numerically unstable. The authors should correct the update rule to a proper FedAvg aggregation (e.g., averaging client updates or using local SGD with a server-side average) or clarify the exact aggregation scheme used in the experiments.
minor comments (6)
  1. [§III-A] The sentence 'By analyzing threats from multiple perspectives.' is an incomplete sentence fragment and should be merged with the preceding or following text.
  2. [Figure 2] The label 'Fusionned' is a typo; it should be 'Fused'.
  3. [Table IV] The header 'False Positive Rate ((%)' contains a doubled opening parenthesis; it should read 'False Positive Rate (%).'
  4. [§III-C (end)] The sentence 'Algorithm 1 outlines the smart contract's on-chain aggregation.' refers to a smart contract and on-chain aggregation, but no smart contract or blockchain component is described anywhere else in the paper. This appears to be a copy-paste error and should be removed or replaced.
  5. [§IV-B, Table III] The paper reports results for 10 clients but does not explain how the CICEVSE2024 dataset is partitioned among the clients (e.g., by station, by time, or randomly). This detail is necessary to interpret the federated results and to assess the effect of non-IID data.
  6. [§V] The conclusion states that the framework provides 'a scalable, secure solution' but the paper does not report training time, communication cost, or any scalability experiments beyond varying the number of clients from 2 to 10. A brief discussion of these costs would strengthen the scalability claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the empirical pipeline is self-contained and does not reduce to its inputs.

full rationale

The paper is an empirical study: features are extracted (Eq. 1), autoencoders compress them (Eq. 2), latent vectors are concatenated (Eq. 3), and a 1D CNN classifies the fused vector (Eqs. 4–6) under federated aggregation (Eq. 7). None of these equations defines an output in terms of the claimed result, nor is any fitted parameter renamed as a prediction. The evaluation uses held-out partitions of the same dataset, which is standard supervised evaluation rather than a fitting-then-predicting cycle. The self-citations in the related work (e.g., [4], [5], [14]–[17]) are surveys of prior network IDS work and do not provide any load-bearing premise for the proposed framework. The unsupported 'outperforms existing solutions' claim and the unresolved sample-alignment in fusion are validity concerns, not circularity. Therefore no circular step can be quoted and exhibited.

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

The central claim rests on the CICEVSE2024 testbed being a faithful model of real EVSE attacks, on the assumption that the two log modalities can be meaningfully paired, and on several manually chosen hyperparameters. No new physical or conceptual entities are introduced; the contribution is purely an empirical ML pipeline.

free parameters (5)
  • Autoencoder bottleneck size = 32
    Chosen by hand to compress each modality; no sensitivity analysis is presented (Section IV-B).
  • Number of clients = 10, 8, 6, 3
    Used in experiments (Table IV) with no motivation for these values.
  • Training epochs = 10
    Fixed for all runs (Table II), with no early stopping or convergence analysis.
  • Batch size = 32
    Fixed hyperparameter (Table II).
  • Learning rate = not reported
    Adam optimizer is used but the learning rate is never stated, so the results are not reproducible.
assumptions (4)
  • domain assumption CICEVSE2024 dataset is representative of real EVSE operating conditions and attacks.
    All experimental conclusions depend on the dataset's fidelity to production EVSE environments (Section IV-A).
  • domain assumption Network and kernel/HPC log samples can be paired or fused as observations of the same station state.
    The fusion step in Eq. (3) concatenates latent vectors from the two modalities, which presumes a one-to-one correspondence across the datasets. The paper never specifies how this pairing is done (Section III-B, Table I).
  • domain assumption Training and test partitions are independent and identically distributed.
    Standard supervised learning assumption, not verified; no split description is provided (Section IV-B).
  • standard math Adam optimizer and categorical cross-entropy loss converge to a good local minimum for this architecture.
    Standard deep learning background, not proven here but widely accepted.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fuse and Federate: Enhancing EV Charging Station Security with Multimodal Fusion and Federated Learning." pith.science (2026). https://pith.science/paper/UJSBZQFY

@misc{pith2026250606730,
  author       = {Pith},
  title        = {Pith review of: Fuse and Federate: Enhancing EV Charging Station Security with Multimodal Fusion and Federated Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UJSBZQFY}},
  note         = {Machine review of arXiv:2506.06730}
}
read the original abstract

The rapid global adoption of electric vehicles (EVs) has established electric vehicle supply equipment (EVSE) as a critical component of smart grid infrastructure. While essential for ensuring reliable energy delivery and accessibility, EVSE systems face significant cybersecurity challenges, including network reconnaissance, backdoor intrusions, and distributed denial-of-service (DDoS) attacks. These emerging threats, driven by the interconnected and autonomous nature of EVSE, require innovative and adaptive security mechanisms that go beyond traditional intrusion detection systems (IDS). Existing approaches, whether network-based or host-based, often fail to detect sophisticated and targeted attacks specifically crafted to exploit new vulnerabilities in EVSE infrastructure. This paper proposes a novel intrusion detection framework that leverages multimodal data sources, including network traffic and kernel events, to identify complex attack patterns. The framework employs a distributed learning approach, enabling collaborative intelligence across EVSE stations while preserving data privacy through federated learning. Experimental results demonstrate that the proposed framework outperforms existing solutions, achieving a detection rate above 98% and a precision rate exceeding 97% in decentralized environments. This solution addresses the evolving challenges of EVSE security, offering a scalable and privacypreserving response to advanced cyber threats

Figures

Figures reproduced from arXiv: 2506.06730 by the authors.

Figure 1
Figure 1. Federated Training of the Intrusion Detection Model [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Comparison of performance metrics for Fusionned, [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Performances comparison between centralized and [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

  1. [1]

    Tracking global data on electric vehicles,

    H. Ritchie and M. Roser, “Tracking global data on electric vehicles,” March 12 2024, accessed: 28/10/2024. [Online]. Available: https://ourworldindata.org/electric-car-sales

    work page 2024

  2. [2]

    V oitures ´electriques : pourquoi les bornes de recharge sont-elles de plus en plus cibl ´ees par les hackers ?

    CNEWS, “V oitures ´electriques : pourquoi les bornes de recharge sont-elles de plus en plus cibl ´ees par les hackers ?”CNEWS, 2024. [Online]. Available: https://www.cnews.fr/france/2024-10-28/voitures- electriques-pourquoi-les-bornes-de-recharge-sont-elles-de-plus-en-plus

    work page 2024

  3. [3]

    Les v ´ehicules ´electriques sont une nouvelle mine d’or pour les hackers,

    Numerama, “Les v ´ehicules ´electriques sont une nouvelle mine d’or pour les hackers,”Numerama, 2023. [Online]. Avail- able: https://www.numerama.com/cyberguerre/1294758-les-vehicules- electriques-sont-une-nouvelle-mine-dor-pour-les-hackers.html

    work page arXiv 2023

  4. [4]

    Beyond detection: Leveraging large language models for cyber attack prediction in iot networks,

    A. Diaf, A. A. Korba, N. E. Karabadji, and Y . Ghamri-Doudane, “Beyond detection: Leveraging large language models for cyber attack prediction in iot networks,” in2024 20th International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT). IEEE, 2024, pp. 117–123

    work page 2024

  5. [5]

    A life-long learning intrusion detection system for 6g-enabled iov,

    A. A. Korba, S. Sebaa, M. Mabrouki, Y . Ghamri-Doudane, and K. Be- natchba, “A life-long learning intrusion detection system for 6g-enabled iov,” in2024 International Wireless Communications and Mobile Com- puting (IWCMC), 2024, pp. 1773–1778

    work page 2024

  6. [6]

    Enhancing anomaly detection in electric vehicle supply equipment (evse) networks using classical and ensemble learning approaches,

    S. R. Hegde, V . V . Chandrakala, G. K. T., and V . K. A. Shankar, “Enhancing anomaly detection in electric vehicle supply equipment (evse) networks using classical and ensemble learning approaches,” in 2024 Control Instrumentation System Conference (CISCON), 2024, pp. 1–5

    work page 2024

  7. [7]

    A novel ensembling of cnn-a-lstm for iot electric vehicle charging stations based on intrusion detection system,

    G. Balakrishna, A. Kumar, T. V ., A. Younas, N. M. G. Kumar, and R. Rastogi, “A novel ensembling of cnn-a-lstm for iot electric vehicle charging stations based on intrusion detection system,” in2023 Inter- national Conference on Self Sustainable Artificial Intelligence Systems (ICSSAS), 2023, pp. 1312–1317

    work page 2023

  8. [8]

    Injection attacks and detection strategy in front-end vehicle-to-grid communication,

    S. Poudel, J. E. Baugh, A. Takiddin, M. Ismail, and S. S. Refaat, “Injection attacks and detection strategy in front-end vehicle-to-grid communication,” in2023 IEEE International Conference on Com- munications, Control, and Computing Technologies for Smart Grids (SmartGridComm), 2023, pp. 1–6

    work page 2023

Show all 24 references
  1. [9]

    Attack design for maximum malware spread through evs commute and charge in power-transportation systems,

    S. Poudel, M. Abouyoussef, J. E. Baugh, and M. Ismail, “Attack design for maximum malware spread through evs commute and charge in power-transportation systems,”IEEE Systems Journal, vol. 18, no. 3, pp. 1809–1820, September 2024

    2024

  2. [10]

    Secure electric vehicle charging infrastructure in smart cities: A blockchain-based smart contract approach,

    A. Chowdhury, S. S. Shafin, S. Masum, J. Kamruzzaman, and D. Shi, “Secure electric vehicle charging infrastructure in smart cities: A blockchain-based smart contract approach,”Smart Cities, vol. 8, no. 1, p. 33, 2025

    2025

  3. [11]

    A blockchain and pki-based secure vehicle-to-vehicle energy-trading protocol,

    M. S. Hossain, C. Rodine, and E. E. Tsiropoulou, “A blockchain and pki-based secure vehicle-to-vehicle energy-trading protocol,”Energies, vol. 17, no. 17, p. 4245, 2024

    2024

  4. [12]

    Deep reinforcement learning-driven miti- gation of adverse effects of cyber-attacks on electric vehicle charging station,

    M. Basnet and M. H. Ali, “Deep reinforcement learning-driven miti- gation of adverse effects of cyber-attacks on electric vehicle charging station,”Energies, vol. 16, no. 21, p. 7296, 2023

    2023

  5. [13]

    Detecting state of charge false reporting attacks via reinforcement learning approach,

    M. A. Alomrani, M. H. K. Tushar, and D. Kundur, “Detecting state of charge false reporting attacks via reinforcement learning approach,” IEEE Transactions on Intelligent Transportation Systems, vol. 24, no. 10, pp. 10 467–10 476, 2023

    2023

  6. [14]

    Ai-driven fast and early detection of iot botnet threats: A comprehensive network traffic analysis approach,

    A. A. Korba, A. Diaf, and Y . Ghamri-Doudane, “Ai-driven fast and early detection of iot botnet threats: A comprehensive network traffic analysis approach,” in2024 International Wireless Communications and Mobile Computing (IWCMC). IEEE, 2024, pp. 1779–1784

    2024

  7. [15]

    Federated learning for zero-day attack detection in 5g and beyond v2x networks,

    A. A. Korba, A. Boualouache, B. Brik, R. Rahal, Y . Ghamri-Doudane, and S. M. Senouci, “Federated learning for zero-day attack detection in 5g and beyond v2x networks,” inICC 2023-IEEE International Conference on Communications. IEEE, 2023, pp. 1137–1142

    2023

  8. [16]

    Multi- agent reinforcement learning-based network intrusion detection system,

    A. Tellache, A. Mokhtari, A. A. Korba, and Y . Ghamri-Doudane, “Multi- agent reinforcement learning-based network intrusion detection system,” inNOMS 2024-2024 IEEE Network Operations and Management Sym- posium. IEEE, 2024, pp. 1–9

    2024

  9. [17]

    Zero-x: A blockchain-enabled open-set federated learning framework for zero-day attack detection in iov,

    A. A. Korba, A. Boualouache, and Y . Ghamri-Doudane, “Zero-x: A blockchain-enabled open-set federated learning framework for zero-day attack detection in iov,”IEEE Transactions on Vehicular Technology, vol. 73, no. 9, pp. 12 399–12 414, 2024

    2024

  10. [18]

    A systematic literature review on host-based intrusion detection systems,

    H. Satilmis ¸, S. Akleylek, and Z. Y . Tok, “A systematic literature review on host-based intrusion detection systems,”IEEE Access, vol. 12, pp. 27 237–27 266, 2024

    2024

  11. [19]

    Private metering in ev charging infrastructure: An ocpp extension,

    N. F. Kamal, A. Sharida, S. Bayhan, H. Alnuweiri, and H. Abu-Rub, “Private metering in ev charging infrastructure: An ocpp extension,” IEEE Transactions on Vehicular Technology, vol. 73, no. 10, pp. 15 456– 15 466, October 2024

    2024

  12. [20]

    Tls-handshake for plug and charge in vehicular communica- tions,

    A. Kilic, “Tls-handshake for plug and charge in vehicular communica- tions,”Computer Networks, vol. 243, p. 110281, 2024

    2024

  13. [21]

    Clustering methodologies for flexibility characterization of electric vehicles supply equipment,

    M. Forte, C. P. Guzman, A. Lekidis, and H. Morais, “Clustering methodologies for flexibility characterization of electric vehicles supply equipment,”Green Energy and Intelligent Transportation, p. 100304, 2025

    2025

  14. [22]

    Transfer learning for securing electric vehicle charging infrastructure from cyber-physical attacks,

    A. Almadhor, S. Alsubai, I. Bouazzi, V . Karovic, M. Davidekova, A. Al Hejaili, and G. A. Sampedro, “Transfer learning for securing electric vehicle charging infrastructure from cyber-physical attacks,” Scientific Reports, vol. 15, no. 1, p. 9331, 2025

    2025

  15. [23]

    A study of ev-evse ecosystem integrity: Machine learning based security monitoring of charging sessions,

    S. Shirvani and A. Ghorbani, “A study of ev-evse ecosystem integrity: Machine learning based security monitoring of charging sessions,” Available at SSRN 4711137, 2024

    2024

  16. [24]

    En- hancing ev charging station security using a multi-dimensional dataset: Cicevse2024,

    E. D. Buedi, A. A. Ghorbani, S. Dadkhah, and R. L. Ferreira, “En- hancing ev charging station security using a multi-dimensional dataset: Cicevse2024,” inData and Applications Security and Privacy XXXVIII, DBSec 2024, Lecture Notes in Computer Science, A. L. Ferrara and R. Kri...

    2024

Pith tools

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