TEE-based Selective Testing of Local Workers in Federated Learning Systems

TL;DR

This paper introduces a TEE-based method for selectively testing local workers in federated learning systems, combining cryptography, smart contracts, and game theory to ensure correctness and security.

Contribution

It presents a novel TEE-based approach that enhances trust and verification in federated learning with theoretical and practical validation.

Findings

The approach is secure against malicious workers.

It is efficient and practical for real-world deployment.

The method outperforms existing verification techniques.

Abstract

This paper considers a federated learning system composed of a central coordinating server and multiple distributed local workers, all having access to trusted execution environments (TEEs). In order to ensure that the untrusted workers correctly perform local learning, we propose a new TEE-based approach that also combines techniques from applied cryptography, smart contract and game theory. Theoretical analysis and implementation-based evaluations show that, the proposed approach is secure, efficient and practical.