Reproduce “Privacy-preserving Federated Learning based on Multi-key Homomorphic Encryption” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jing Ma, Si-Ahmed Naas, Stephan Sigg, Xixiang Lyu
A privacy-preserving federated learning scheme is introduced based on xMK-CKKS, an enhanced multi-key homomorphic encryption protocol, to secure model updates in IoT environments. This approach maintains model accuracy comparable to traditional federated learning while significantly reducing computational overhead for resource-constrained devices and demonstrating resilience against collusion attacks.
Key insights
- Aggregating public keys simplifies the encryption process and distributes trust, making decryption dependent on all participants.
- Designing the decryption process to require collective participation (decryption shares) can effectively prevent collusion attacks between the server and a subset of clients.
- Tailoring multi-key homomorphic encryption schemes (like CKKS) can achieve both privacy and computational efficiency suitable for resource-constrained IoT devices in federated learning.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.