Efficient Privacy-Preserving Cross-Silo Federated Learning with Multi-Key Homomorphic Encryption

TL;DR

MASER is an efficient privacy-preserving federated learning framework using multi-key homomorphic encryption, significantly reducing computation and communication costs while maintaining accuracy.

Contribution

It introduces MASER, a novel MKHE-based FL framework that employs consensus-based pruning and slicing to improve efficiency for real-world deployment.

Findings

MASER is 3.03 to 8.29 times more efficient than existing approaches.

Maintains comparable accuracy to standard FL algorithms.

Overhead is only 1.48 to 5 times higher than vanilla FL.

Abstract

Federated Learning (FL) is susceptible to privacy attacks, such as data reconstruction attacks, in which a semi-honest server or a malicious client infers information about other clients' datasets from their model updates or gradients. To enhance the privacy of FL, recent studies combined Multi-Key Homomorphic Encryption (MKHE) and FL, making it possible to aggregate the encrypted model updates using different keys without having to decrypt them. Despite the privacy guarantees of MKHE, existing approaches are not well-suited for real-world deployment due to their high computation and communication overhead. We propose MASER, an efficient MKHE-based Privacy-Preserving FL framework that combines consensus-based model pruning and slicing techniques to reduce this overhead. Our experimental results show that MASER is 3.03 to 8.29 times more efficient than existing MKHE-based FL approaches in terms of computation and communication overhead while maintaining comparable classification accuracy to standard FL algorithms. Compared to a vanilla FL algorithm, the overhead of MASER is only 1.48 to 5 times higher, striking a good balance between privacy, accuracy, and efficiency in both IID and non-IID settings.