Authors : Riya Patel; Santosh Saha

Volume/Issue : Volume 11 - 2026, Issue 6 - June

Google Scholar : https://tinyurl.com/phm7txyz

Scribd : https://tinyurl.com/mvw44tn8

DOI : https://doi.org/10.38124/ijisrt/26jun570

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : The current increase in cyber threats highlights the need for an advanced intrusion detection system.While IoT networks have be-come more vulnerable to various types of network-level attacks, most AIpowered IDSs are centralized, black-box, and susceptible to manipu-lations of their models’ parameters. This work proposes XAI-FedGuard, a federated learning approach which addresses all three issues simultane-ously. The edge devices locally train a lightweight CNN-LSTM classifier based on traffic data, so there is no need to send raw traffic informa-tion to the server. A SHA-256 hash-chain mechanism ensures integrity by checking whether any model updates can possibly be used in poi-soning attacks regardless of their degree of subtlety. Upon global model convergence, KernelSHAP explains every classification decision based on certain network features.

Keywords : IoT Security · Federated Learning · Intrusion Detection · Explainable AI · Model Integrity.

References :

1. Saheed, Y.K., et al.: Machine learning-based intrusion detection for IoT network attacks. Alexandria Engineering Journal 61, 9395–9409 (2022)
2. Karimullah, K., et al.: A hybrid CNN-LSTM-based intrusion detection system trained on UNSW-NB15. Journal of Computing and Biomedical Informatics 10(1) (2025)
3. Mallidi, S.K.R., Ramisetty, R.R.: Advancements in AI-based intrusion detection systems in IoT: a review. Discover Internet of Things 5, 8 (2025)
4. McMahan, B., et al.: Communication-efficient learning of deep networks from de-centralized data. In: Proceedings of AISTATS (2017)
5. Latif, S., et al.: Lightweight integrity-driven federated learning for IoT security. IEEE Internet of Things Journal (2024)
6. Lundberg, S.M., Lee, S.I.: A unified approach to interpreting model predictions. In: Advances in Neural Information Processing Systems (NeurIPS) (2017)
7. Ribeiro, M.T., et al.: ‘Why should I trust you?’: explaining the predictions of any classifier. In: Proceedings of KDD (2016)
8. Fatema, N., et al.: FedXAI-IDS: federated explainable intrusion detection for IoT. IEEE Access (2025)
9. Bhattacharjee, S., et al.: A probabilistic trust framework for secure IoT transmis-sions. IEEE Sensors Journal (2017)
10. Reis, A.: Hybrid deep learning for anomaly detection in 5G-enabled smart city IoT. Sensors 22(21), 8417 (2022)
11. Dirin, A., et al.: IoTAttest: TPM 2.0 remote attestation for IoT device identity. IEEE Access (2023)
12. Koppula, V., Leo Joseph, J.: IoT Vulnerability Dataset. Mendeley Data v1 (2024). https://doi.org/10.17632/7m58kxs742.1
13. Alrayes, F.S., et al.: Optimizing intrusion detection using hybrid random forest and CNN-LSTM. Journal of AI Research (2025)
14. Bhavsar, M., et al.: Anomaly-based intrusion detection system for IoT applications. Discover IoT 3, 5 (2023)
15. Alkhonaini, M.A., et al.: A two-phase spatiotemporal chaos-based protocol for IoT data integrity. Scientific Reports 14(1) (2024)
16. Hang, I., Kim, D.: A study of IoT security and blockchain. Sensors 19(7), 1729 (2019)
17. Zhao, G., et al.: Privacy-preserving blockchain-based integrity checking for IoT. IEEE Transactions on Cloud Computing (2020)
18. Aman, M.N., et al.: Low power data integrity in IoT systems. IEEE Internet of Things Journal 5(4), 3102–3113 (2018)

19. Li, X., Huang, K., Yang, W., Wang, S., Zhang, Z.: On the convergence of FedAvg on non-IID data. In: International Conference on Learning Representations (2020). https://openreview.net/forum?id=HJxNAnVtDS