On the Push-Based Asynchronous Federated Learning: A Bias-Correction Aggregation Approach

2026-05-24

AI Summary

This work addresses the challenges of aggregation bias, model drift, and high communication overhead in asynchronous decentralized federated learning caused by non-IID data, directed network topologies, and frequent communication. To this end, the authors propose PushCen-ADFL, a novel framework that integrates communication, aggregation, and local stabilization within a shared centroid representation space. The method employs a mean-preserving push-sum hybrid protocol to correct aggregation bias, introduces lightweight centroid regularization to mitigate model drift induced by data heterogeneity and asynchrony, and incorporates a sender-side deduplication buffer to enhance robustness under asynchronous conditions. Experimental results demonstrate that PushCen-ADFL achieves up to a 6% improvement in accuracy on vision benchmarks while reducing per-communication cost by over 80%, substantially improving the trade-off between model accuracy and communication efficiency.

Abstract

Asynchronous decentralized federated learning (ADFL) eliminates central coordination and global synchronization, making it attractive for large-scale and heterogeneous systems. However, frequent peer-to-peer communication, asynchronous updates on directed topologies, and non-IID data jointly lead to excessive communication overhead, biased aggregation and severe model drift. We propose PushCen-ADFL, a communication-efficient ADFL framework that enables stable training under asymmetric communication and delayed client participation. PushCen-ADFL couples communication, aggregation, and local stabilization in a shared centroid representation space, forming a closed loop between compression and optimization. Clients exchange centroid-form messages, apply average-preserving push-sum mixing to correct aggregation bias, and use a lightweight centroid regularization anchored in the same centroid space to mitigate drift under heterogeneity and staleness. A bounded, sender-deduplicated buffer further improves robustness under irregular asynchronous arrivals. Experiments on vision datasets demonstrate that PushCen-ADFL improves accuracy under data heterogeneity by up to 6\% while reducing per-push communication cost by more than 80\%, achieving a favorable accuracy-communication trade-off.

Problem

• asynchronous federated learning
• decentralized learning
• aggregation bias
• model drift
• non-IID data

Innovation

• asynchronous federated learning
• bias correction
• centroid representation
• push-sum mixing
• communication efficiency