Route-and-Aggregate Decentralized Federated Learning Under Communication Errors

2025-03-28

AI Summary

To address the high flooding overhead and poor robustness of conventional gossip-based decentralized federated learning (D-FL) under unreliable communication, this paper proposes a Routing-and-Aggregation (R&A) D-FL framework. It enables directed model exchange—rather than network-wide flooding—via end-to-end packet error rate–driven optimal routing selection. Furthermore, it jointly models routing paths and aggregation weights, introducing an error-compensating adaptive normalized weighting mechanism. This work establishes, for the first time, a theoretical coupling between D-FL convergence and network-layer routing characteristics. Experiments demonstrate that, on a 10-client network, R&A-D-FL achieves 35% higher accuracy than standard gossip D-FL. When scaled to a 28-node network with bit errors, its performance approaches that of error-free centralized FL. The framework significantly improves both convergence speed and robustness in heterogeneous, lossy networks.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Distributed Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Decentralized federated learning (D-FL) allows clients to aggregate learning models locally, offering flexibility and scalability. Existing D-FL methods use gossip protocols, which are inefficient when not all nodes in the network are D-FL clients. This paper puts forth a new D-FL strategy, termed Route-and-Aggregate (R&A) D-FL, where participating clients exchange models with their peers through established routes (as opposed to flooding) and adaptively normalize their aggregation coefficients to compensate for communication errors. The impact of routing and imperfect links on the convergence of R&A D-FL is analyzed, revealing that convergence is minimized when routes with the minimum end-to-end packet error rates are employed to deliver models. Our analysis is experimentally validated through three image classification tasks and two next-word prediction tasks, utilizing widely recognized datasets and models. R&A D-FL outperforms the flooding-based D-FL method in terms of training accuracy by 35% in our tested 10-client network, and shows strong synergy between D-FL and networking. In another test with 10 D-FL clients, the training accuracy of R&A D-FL with communication errors approaches that of the ideal C-FL without communication errors, as the number of routing nodes (i.e., nodes that do not participate in the training of D-FL) rises to 28.

Problem

• Improving decentralized federated learning efficiency under communication errors
• Optimizing model aggregation with adaptive routing and normalization
• Enhancing training accuracy in non-ideal network conditions

Innovation

• Route-and-Aggregate decentralized federated learning strategy
• Adaptive normalization for aggregation coefficients
• Optimized routing minimizes convergence impact