DAG-ACFL: Asynchronous Clustered Federated Learning based on DAG-DLT

TL;DR

DAG-ACFL introduces an asynchronous, decentralized federated learning framework utilizing DAG-DLT, with novel tip selection algorithms to improve clustering and training efficiency while reducing communication and storage costs.

Contribution

It presents DAG-ACFL, a new asynchronous clustered federated learning framework based on DAG-DLT, with innovative tip selection algorithms for improved model aggregation.

Findings

DAG-ACFL outperforms existing frameworks in asynchronous clustered FL.

The proposed algorithms effectively cluster clients with similar data distributions.

DAG-ACFL demonstrates reduced communication and storage costs.

Abstract

Federated learning (FL) aims to collaboratively train a global model while ensuring client data privacy. However, FL faces challenges from the non-IID data distribution among clients. Clustered FL (CFL) has emerged as a promising solution, but most existing CFL frameworks adopt synchronous frameworks lacking asynchrony. An asynchronous CFL framework called SDAGFL based on directed acyclic graph distributed ledger techniques (DAG-DLT) was proposed, but its complete decentralization leads to high communication and storage costs. We propose DAG-ACFL, an asynchronous clustered FL framework based on directed acyclic graph distributed ledger techniques (DAG-DLT). We first detail the components of DAG-ACFL. A tip selection algorithm based on the cosine similarity of model parameters is then designed to aggregate models from clients with similar distributions. An adaptive tip selection algorithm leveraging change-point detection dynamically determines the number of selected tips. We evaluate the clustering and training performance of DAG-ACFL on multiple datasets and analyze its communication and storage costs. Experiments show the superiority of DAG-ACFL in asynchronous clustered FL. By combining DAG-DLT with clustered FL, DAG-ACFL realizes robust, decentralized and private model training with efficient performance.