LEO satellites speed up federated learning over large remote areas
ParaBlock++: Communication-Computation Parallel Blockwise Hierarchical Federated Learning
Machine Learning
Summary
Training AI models on devices spread across large or remote areas is slow and hard due to limited resources and connections. The authors study a system called ParaBlock++ that uses low Earth orbit satellites to help devices train AI models together faster by sharing updates more efficiently. The satellites not only connect devices but also help with training themselves and relay information quickly between each other. Their method uses a new way of updating models that converges faster and lowers communication costs. Tests show this approach reaches good AI accuracy faster with less overall communication.
What this means in practice
- •For satellite network operators: Reduce latency and communication in federated learning tasks by using LEO satellites for model aggregation and relay in wide-area ML deployments.
- •For mobile edge computing teams: Enable collaborative AI training with limited local computing power by integrating satellite-assisted hierarchical federated learning.
Authors
Kai Wang, Chee Wei Tan
Abstract
In the emerging 6G paradigm, where AI-driven services operate over decentralized and resource-constrained infrastructures, the convergence of distributed learning algorithms remains a significant challenge. This paper introduces $\mathrm{ParaBlock}$++, a low Earth orbit (LEO) satellite-assisted hierarchical federated learning (HFL) framework, enabling resource-limited terrestrial clients to collaboratively engage in federated learning (FL) via ground-satellite links (GSLs) supported by LEO satellite megaconstellations. Within this architecture, LEO satellites serve multiple roles: as edge servers aggregating and broadcasting models, as clients accelerating training, and as high-throughput relays managing inter-satellite routing through inter-satellite links (ISLs). Departing from conventional synchronous Jacobian-type updates, $\mathrm{ParaBlock}$++ employs asynchronous BCD-type updates to enhance convergence speed. We provide theoretical convergence guarantees for convex, strongly convex, and non-convex objectives, and formulate optimization problems to determine optimal cluster resource allocation and model partitioning for minimizing global training latency. Experimental evaluations demonstrate that $\mathrm{ParaBlock}$++ achieves faster initial convergence and comparable final performance for deep neural networks (DNNs) such as ResNet-18, offering promising directions for federated learning across geographically remote areas enabled by 6G technologies, but much less global rounds and communication cost.