Towards a Distributed Federated Learning Aggregation Placement using Particle Swarm Intelligence

2025-04-22

AI Summary

This work addresses the dynamic aggregator placement problem in hierarchical semi-decentralized federated learning (SDFL). To reduce system dependency and minimize end-to-end latency, we propose a latency-driven, adaptive aggregator placement method that avoids real-time resource monitoring. Differing from conventional approaches, our method is the first to employ Particle Swarm Optimization (PSO) for aggregator location selection—relying solely on an end-to-end processing latency model—thereby significantly reducing communication and state-synchronization overhead. Leveraging Docker-based containerization and integrating seamlessly with hierarchical FL architecture, the method enables efficient, lightweight deployment in realistic frameworks. Experimental results demonstrate that, compared to random and uniform placement strategies, our approach reduces total processing time by 43% and 32%, respectively, while maintaining rapid convergence under large-scale client settings—validating its scalability and practical applicability.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Search and Optimization: Distributed Search
• Multiagent Systems: Distributed Problem Solving

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Federated learning has become a promising distributed learning concept with extra insurance on data privacy. Extensive studies on various models of Federated learning have been done since the coinage of its term. One of the important derivatives of federated learning is hierarchical semi-decentralized federated learning, which distributes the load of the aggregation task over multiple nodes and parallelizes the aggregation workload at the breadth of each level of the hierarchy. Various methods have also been proposed to perform inter-cluster and intra-cluster aggregation optimally. Most of the solutions, nonetheless, require monitoring the nodes' performance and resource consumption at each round, which necessitates frequently exchanging systematic data. To optimally perform distributed aggregation in SDFL with minimal reliance on systematic data, we propose Flag-Swap, a Particle Swarm Optimization (PSO) method that optimizes the aggregation placement according only to the processing delay. Our simulation results show that PSO-based placement can find the optimal placement relatively fast, even in scenarios with many clients as candidates for aggregation. Our real-world docker-based implementation of Flag-Swap over the recently emerged FL framework shows superior performance compared to black-box-based deterministic placement strategies, with about 43% minutes faster than random placement, and 32% minutes faster than uniform placement, in terms of total processing time.

Problem

• Optimize aggregation placement in hierarchical federated learning
• Minimize reliance on systematic data during aggregation
• Improve processing delay in distributed federated learning

Innovation

• Uses Particle Swarm Optimization for aggregation placement
• Minimizes reliance on systematic data monitoring
• Optimizes processing delay in distributed aggregation