DFL: High-Performance Blockchain-Based Federated Learning

TL;DR

This paper introduces DFL, a blockchain architecture optimized for federated learning that achieves high accuracy and robustness with low resource consumption by eliminating global consensus.

Contribution

The paper proposes DFL, a novel blockchain design that reduces latency and resource use, enabling efficient and secure federated learning without global consensus.

Findings

Achieves over 90% accuracy on non-IID datasets

Maintains robustness against model poisoning attacks

Uses less than 5% of hardware resources

Abstract

Many researchers have proposed replacing the aggregation server in federated learning with a blockchain system to improve privacy, robustness, and scalability. In this approach, clients would upload their updated models to the blockchain ledger and use a smart contract to perform model averaging. However, the significant delay and limited computational capabilities of blockchain systems make it inefficient to support machine learning applications on the blockchain. In this paper, we propose a new public blockchain architecture called DFL, which is specially optimized for distributed federated machine learning. Our architecture inherits the merits of traditional blockchain systems while achieving low latency and low resource consumption by waiving global consensus. To evaluate the performance and robustness of our architecture, we implemented a prototype and tested it on a physical four-node network, and also developed a simulator to simulate larger networks and more complex situations. Our experiments show that the DFL architecture can reach over 90\% accuracy for non-I.I.D. datasets, even in the presence of model poisoning attacks, while ensuring that the blockchain part consumes less than 5\% of hardware resources.