SuperSFL: Resource-Heterogeneous Federated Split Learning with Weight-Sharing Super-Networks

2026-01-05

arXiv.org

AI Summary

This work proposes SuperSFL, a novel framework addressing the inefficiencies of split federated learning (SFL) on heterogeneous edge devices, including slow convergence, high communication overhead, and low training efficiency. SuperSFL leverages a weight-sharing hypernetwork to dynamically generate resource-adaptive subnetworks for diverse clients and introduces a three-phase gradient fusion (TPGF) mechanism that jointly optimizes local updates, server-side computation, and gradient aggregation. To enhance robustness under communication failures, it further integrates a fault-tolerant client classifier with a collaborative aggregation strategy. Experimental results on CIFAR-10 and CIFAR-100 demonstrate that SuperSFL reduces communication rounds by 2–5×, cuts total communication cost by up to 20×, shortens training time by 13×, and significantly improves both accuracy and energy efficiency compared to baseline SFL approaches.

Technology Category

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

Application Category

• 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: Efficiency and scalability of Web search engines
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

SplitFed Learning (SFL) combines federated learning and split learning to enable collaborative training across distributed edge devices; however, it faces significant challenges in heterogeneous environments with diverse computational and communication capabilities. This paper proposes \textit{SuperSFL}, a federated split learning framework that leverages a weight-sharing super-network to dynamically generate resource-aware client-specific subnetworks, effectively mitigating device heterogeneity. SuperSFL introduces Three-Phase Gradient Fusion (TPGF), an optimization mechanism that coordinates local updates, server-side computation, and gradient fusion to accelerate convergence. In addition, a fault-tolerant client-side classifier and collaborative client--server aggregation enable uninterrupted training under intermittent communication failures. Experimental results on CIFAR-10 and CIFAR-100 with up to 100 heterogeneous clients show that SuperSFL converges $2$--$5\times$ faster in terms of communication rounds than baseline SFL while achieving higher accuracy, resulting in up to $20\times$ lower total communication cost and $13\times$ shorter training time. SuperSFL also demonstrates improved energy efficiency compared to baseline methods, making it a practical solution for federated learning in heterogeneous edge environments.

Problem

• Federated Learning
• Split Learning
• Device Heterogeneity
• Edge Computing
• Resource Heterogeneity

Innovation

• SuperSFL
• weight-sharing super-network
• Three-Phase Gradient Fusion
• resource heterogeneity
• federated split learning