Hybrid Homomorphic Encryption-Differential Privacy for Resource-Constrained IoT

Khan et al. (2025) and Mahmood et al. (2025) compare HE and DP in isolation, but IoT devices (per Robai, 2024) often lack resources for either. This research proposes a hybrid protocol where edge devices with sufficient compute use HE for high-sensitivity data (e.g., medical readings), while resource-constrained devices apply DP with calibrated noise. A central coordinator would partition tasks dynamically, inspired by HySec-Flow's (Widanage et al., 2021) container-based workload distribution. Unlike Rezak et al.'s (2023) trade-off analysis, this system optimizes the trade-off by contextually selecting techniques. For example, in ToN-IoT networks (per Mahmood et al., 2025), critical alerts would use HE, while routine telemetry uses DP. The innovation lies in adaptive privacy technique selection to balance security, accuracy, and efficiency.

References:

1. Federated Learning with Privacy-Preserving Big Data Analytics for Distributed Healthcare Systems. Shuchona Malek Orthi, Md Habibur Rahman, ✉. Kazi, Bushra Siddiqa, Mukther Uddin, Sazzat Hossain, Mohd Abdullah Al Mamun, Nazibullah Khan (2025). Journal of Computer Science and Technology Studies.
2. Privacy Preserving Analytics in IoT Systems using Federated Learning with Homomorphic Encryption and Differential Privacy. Abuthar Mahmood, Sasikumar Gurumoorthy, Meghana A, Yogesh Ramaswamy, Gayathiri B (2025). 2025 4th International Conference on Distributed Computing and Electrical Circuits and Electronics (ICDCECE).
3. Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy. Fawad Khan, Syed Yaseen Shah, Syed Aziz Shah, Jawad Ahmad, Shahzaib Tahir, Adnan Zahid (2025). 2025 International Telecommunications Conference (ITC-Egypt).
4. Exploring Homomorphic Encryption and Differential Privacy Techniques towards Secure Federated Learning Paradigm. Rezak Aziz, S. Banerjee, S. Bouzefrane, Thinh Le Vinh (2023). Future Internet.
5. Federated learning for secure and privacy preserving data analytics in heterogeneous networks. Mmasi Patience Robai, Mmasi Patience, Robai (2024). GSC Advanced Research and Reviews.
6. HySec-Flow: Privacy-Preserving Genomic Computing with SGX-based Big-Data Analytics Framework. Chathura Widanage, Weijie Liu, Jiayu Li, Hongbo Chen, Xiaofeng Wang, Haixu Tang, Judy Fox (2021). IEEE International Conference on Cloud Computing.