A Low-Latency Secure Federated Learning Framework for IoT Environments Using VPN Mesh and Homomorphic Encryption over Decentralized SDS
DOI:
https://doi.org/10.19139/soic-2310-5070-4144Keywords:
Federated Learning, Federated Edge Learning, Decentralized Architecture, Software Defined Storage, VPNAbstract
Given the continuous evolution of cyberattacks in digital environments and the widespread adoption of intelligent technologies, cybersecurity challenges have significantly increased worldwide. These challenges are particularly critical in Federated Learning environments, which require efficient networks for training high-performing models. Advanced Persistent Threats further increase these risks by targeting servers and Internet of Things devices through sophisticated and long-term attack strategies. Consequently, the management and security of IoT infrastructures in Federated environments become more complex. Such attacks can disrupt the training process, delay model convergence, compromise parameter integrity, and produce biased or poisoned updates, resulting in unreliable models. To address these challenges, this paper proposes a Framework designed to secure learning parameters during the training phase in Federated Learning environments while reducing latency, along with an algorithm to improve the efficiency of the model training process. This Framework contributes to protecting the federated environment against Data Poisoning, Model Inversion, and Integrity Attacks through the implementation of a decentralized SDS architecture combined with a VPN-secured mesh topology. Experimental results show that the proposed Framework achieves an accuracy rate of 98.58%, while maintaining a latency of 31.04 under non-IID data constraints.Downloads
Published
2026-11-16
How to Cite
MAHMOUDI, S., Amnai, M., & El Moudden, T. (2026). A Low-Latency Secure Federated Learning Framework for IoT Environments Using VPN Mesh and Homomorphic Encryption over Decentralized SDS. Statistics, Optimization & Information Computing. https://doi.org/10.19139/soic-2310-5070-4144
License
Copyright (c) 2026 Saad MAHMOUDI, Mohamed Amnai, Tarik El Moudden

This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).