Verifiable Secure Aggregation Based on Functional Encryption for Federated Learning on IoT Devices
IEEE Internet of Things Journal · 15.06.2026 · 13 (12) · 27776-27789
DOI: 10.1109/jiot.2026.3680237 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Internet of Things Journal, erschienen am 15.06.2026. Verfasst von Wei Wu und 4 weiteren Autor:innen. Die Arbeit nennt 39 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Federated learning (FL) enables collaborative model training across multiple Internet of Things (IoT) devices while preserving data privacy. However, the trustworthiness of the aggregation server (AS) remains a critical security vulnerability, especially in IoT environments that rely on potentially untrusted servers. Existing secure aggregation schemes exhibit critical flaws: verification mechanisms and aggregation processes are decoupled, allowing a malicious server to generate valid verification tokens while returning incorrect aggregation results. This enables covert attacks where verification succeeds despite erroneous models, thereby seriously compromising system reliability. As a result, designing a privacy-preserving aggregation scheme that integrates verification with aggregation, while keeping both low computational and communication costs, remains a persistent challenge. To address this issue, we propose a lightweight verifiable secure aggregation based on functional encryption for federated learning on IoT devices (VFE). Our protocol cryptographically binds multiclient function encryption to identity-based aggregate signatures, thereby deeply integrating aggregation computation with verification. We further employ an efficient bilinear-pairing-based verification protocol that supports single-step verification, thereby eliminating auxiliary mechanisms and significantly reducing verification complexity. The security analysis and extensive testing demonstrate that the proposed VFE protocol achieves robust verifiable aggregation in FL, while simultaneously preserving client privacy and substantially reducing both computation and communication overhead, making it highly suitable for IoT deployments.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Wu, W., Zhang, A., Li, H., Hu, F., & Jiang, Y. (2026). Verifiable Secure Aggregation Based on Functional Encryption for Federated Learning on IoT Devices. IEEE Internet of Things Journal, 13(12), 27776-27789. https://doi.org/10.1109/jiot.2026.3680237
Vancouver
Wu W, Zhang A, Li H, Hu F, Jiang Y. Verifiable Secure Aggregation Based on Functional Encryption for Federated Learning on IoT Devices. IEEE Internet of Things Journal. 2026;13(12):27776-27789. doi:10.1109/jiot.2026.3680237
BibTeX
@article{Wu2026,
author = {Wu, Wei and Zhang, Aiqing and Li, Heju and Hu, Fangjie and Jiang, Yili},
title = {Verifiable Secure Aggregation Based on Functional Encryption for Federated Learning on IoT Devices},
journal = {IEEE Internet of Things Journal},
year = {2026},
volume = {13},
number = {12},
pages = {27776--27789},
doi = {10.1109/jiot.2026.3680237}
}
EUVIMED