Composable Lightweight MPC Primitives for Ultra-Constrained IoT Devices

Many IoT authentication protocols (Cetintav & Sandikkaya, 2025; Youssef et al., 2020) focus on pairwise authentication or simple data aggregation, but rarely on general-purpose MPC due to resource constraints. However, the growing need for privacy-preserving sensor fusion, federated learning, and anomaly detection at the edge calls for more capable yet efficient MPC. This research would design new arithmetization-oriented ciphers (see Aly et al., 2020) and lightweight secret-sharing or garbling techniques that minimize memory accesses, communication rounds, and energy use. The primitives would be composable, supporting a range of tasks (e.g., secure sensor aggregation, threshold-based alerts, lightweight private inference). Such a toolkit could make privacy-preserving computation feasible for vast IoT deployments—e.g., in smart cities or healthcare—where current MPC approaches remain too heavyweight.

References:

1. A Review of Lightweight IoT Authentication Protocols From the Perspective of Security Requirements, Computation, Communication, and Hardware Costs. Isil Cetintav, Mehmet Tahir Sandikkaya (2025). IEEE Access.
2. Hardware Implementation of Secure Lightweight Cryptographic Designs for IoT Applications. W. H. Youssef, Ali Abdelli, Fethi Dridi, Mohsen Machhout (2020). Secur. Commun. Networks.
3. Design of Symmetric-Key Primitives for Advanced Cryptographic Protocols. A. Aly, T. Ashur, Eli Ben-Sasson, Siemen Dhooghe, Alan Szepieniec (2020). IACR Transactions on Symmetric Cryptology.