TARU PUBLICATIONS
Journal of Discrete Mathematical Sciences and Cryptography cover

Journal of Discrete Mathematical Sciences and Cryptography

(Since 1998)
Hybrid ·Peer-reviewed·ISSN (Online): 2169-0065·ISSN (Print): 0972-0529

CiteScore 2025: 2.7 (Q1)

Powered by:Powered by

Monthly Journal: Publishes theoretical and applied research in all areas of Discrete Mathematical Sciences, Cryptography, Combinatorics, Elliptic Curves and Information Security.

Issues up to 2022 co-published with and available at:Taylor & Francis Online
submissions@tarupublications.com
Taru Online First Current IssueSpecial IssuesBrowse all Issues
Journal of Discrete Mathematical Sciences and Cryptography
Open Access Research Article

CyberShield-H : A hybrid sequential deep learning model for robust DDoS mitigation in critical healthcare IoT systems

* ,

* Corresponding author · click or hover a name for details

pp. 2889–2898Vol. 28Issue 7October 2025DOI: 10.47974/JDMSC-2559 Crossmark XML
Received:
08 Jul 2025
Published Online:
31 Oct 2025
Article type:
Research Article
Language:
EN
Article no.:
JDMSC-2559
Pages:
2889–2898

Abstract

The growing reliance on Healthcare IoT (H-IoT) systems creates the potential for Distributed Denial of Service (DDoS) attacks on critical services. This research introduces a hybrid deep learning model with a combination of GRU and LSTM networks for real-time, multi-class DDoS detection from flow-aware temporal traffic consecutive data. To enhance robustness, adversarial training with the Fast Gradient Sign Method (FGSM) was used with PCA for efficient operational efficiency with less dimensionality while retaining discriminative features. The adversarially trained GRU-LSTM model achieved a clean data accuracy level of 94.99% and an adversarial accuracy level of 91.17% which is a good measure of robustness. Additionally, interpretability was articulated and accomplished using LIME to show the contributions of the features used in the model. Overall, this work introduces an interpretable and stable framework to provide an insight into the security implications and considerations for H-IoT facing demonstrated DDoS attacks.

Keywords

Adversarial trainingDeep learningDDoS detectionHealthcare IoTLIME interpretabilityTemporal sequence modeling

Subject Classifications

Primary 68T07Secondary 68M25

References

[1] R. Udayakumar, A4. Joshi, S. and R. Sugumar, Deep Fraud Net: A Deep Learning Approach for Cyber Security and Financial Fraud Detection and Classification, Journal of Information Systems and Information Security, vol. 13, no. 4, pp. 138–157, (2023).
[2] DDoS Attacks in Healthcare, Center for Internet Security, (2025).
[3] M. Mittal and S. Behal, Deep learning approaches for detecting DDoS attacks - a systematic review, Soft Computing, vol. 27, no. 1, pp. 1–37, (2021).
[4] CIC- DDoS2019 dataset, Kaggle, (2025).
[5] A. K. Mousa and M. N. Abdullah, An Improved Deep Learning Model for DDoS Detection Based on Hybrid Stacked Autoencoder and Checkpoint Network, Future Internet, vol. 15, no. 8, pp. 278, (2023).
[6] F. Martinez, M. Mapkar, and A. Alfatemi, Redefining DDoS Attack Detection Using a Dual-Space Prototypical Network-Based Approach, Computer Communications and Networks, pp. 279–287, (2024).
[7] M. S. Elsayed, N.-A. Le-Khac, S. Dev, and A. D. Jurcut, DDoSNet: A Deep-Learning Model for Detecting Network Attacks, Computers & Security, vol. 89, pp. 101-129, (2020).
[8] I. Bolodurina, A. Shukhman, and D. Parfenov, Classifying Unbalanced Datasets in Identifying DDoS Attacks, Journal of Physics: Conference Series, vol. 1679, no. 4, pp. 042020, (2020).
[9] J. Mante and K. Kolhe, A novel DDoS detection framework leveraging hybrid binary grey wolf optimization and explainable AI, Journal of Information & Optimization Sciences, vol. 46, no. 2, pp. 563–578, (2025).
[10] P. K. Swain, A. Patnaik, and S. Satpathy, Hybrid machine learning model for network anamoly detection using time-series forecasting, Journal of Information & Optimization Sciences, vol. 46, no. 6, pp. 2311-2322, (2025).

Views: 111Downloads: 80Citations: 0