mlsanityData to 7 Oct 2026

GAPSL: A Gradient-Aligned Parallel Split Learning over Data-Heterogeneous Edge Computing Systems

arXiv 2603.18540 ↗First posted 19 Mar 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

The increasing complexity of neural networks poses significant challenges for democratizing federated learning (FL) on resource-constrained edge devices. Parallel split learning (PSL) has emerged as a promising solution by offloading substantial computing workload to a server via model partitioning, shrinking client-side computing load, and eliminating the client-side model aggregation for reduced communication and deployment costs. However, the highly heterogeneous nature of client data in edge computing systems causes aggregation-free PSL to suffer from severe training divergence, stemming from gradient directional inconsistency across clients. To address this challenge, we propose GAPSL, a gradient-aligned PSL framework tailored for data-heterogeneous edge systems, which comprises two key components: leader gradient identification (LGI) and gradient direction alignment (GDA). LGI dynamically selects a set of directionally consistent device gradients to construct a leader gradient as a robust proxy for the global convergence trend. GDA employs a direction-aware regularization to align each client's gradient with the leader gradient, thereby mitigating inter-device gradient directional inconsistency and enhancing model convergence. We evaluate GAPSL on a prototype computing testbed. Extensive experiments demonstrate that GAPSL consistently outperforms state-of-the-art benchmarks in training accuracy, convergence latency, and system robustness under severe data heterogeneity.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 34 of its 49 references that are in the map. First posted on arXiv 19 Mar 2026.

Builds on

34 of 49 references.

References · 34

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
  4. 2018Split learning for health: Distributed deep learning without sharing raw patient datacited by 391
  5. 2020SplitFed: When Federated Learning Meets Split Learningcited by 308
  6. 2020Unleashing the Tiger: Inference Attacks on Split Learningcited by 104
  7. 2023Efficient Parallel Split Learning over Resource-constrained Wireless Edge Networkscited by 95
  8. 2022Split Learning over Wireless Networks: Parallel Design and Resource Managementcited by 74
  9. 2023Split Learning in 6G Edge Networkscited by 68
  10. 2024AdaptSFL: Adaptive Split Federated Learning in Resource-constrained Edge Networkscited by 64
  11. 2024Automated Federated Pipeline for Parameter-Efficient Fine-Tuning of Large Language Modelscited by 49
  12. 2023FedSN: A Federated Learning Framework over Heterogeneous LEO Satellite Networkscited by 38
22 more
  1. 2024Hierarchical Split Federated Learning: Convergence Analysis and System Optimizationcited by 38
  2. 2023Towards Personalized Federated Learning via Heterogeneous Model Reassemblycited by 31
  3. 2025HSplitLoRA: A Heterogeneous Split Parameter-Efficient Fine-Tuning Framework for Large Language Modelscited by 30
  4. 2024Unleashing the Power of Continual Learning on Non-Centralized Devices: A Surveycited by 23
  5. 2021Server-Side Local Gradient Averaging and Learning Rate Acceleration for Scalable Split Learningcited by 22
  6. 2022A Bargaining Game for Personalized, Energy Efficient Split Learning over Wireless Networkscited by 22
  7. 2025LEO-Split: A Semi-Supervised Split Learning Framework over LEO Satellite Networkscited by 22
  8. 2025FedAWA: Adaptive Optimization of Aggregation Weights in Federated Learning Using Client Vectorscited by 19
  9. 2023FedCache: A Knowledge Cache-driven Federated Learning Architecture for Personalized Edge Intelligencecited by 17
  10. 2025HASFL: Heterogeneity-aware Split Federated Learning over Edge Computing Systemscited by 15
  11. 2024Multi-level Personalized Federated Learning on Heterogeneous and Long-Tailed Datacited by 14
  12. 2023FedCiR: Client-Invariant Representation Learning for Federated Non-IID Featurescited by 12
  13. 2021Splitfed learning without client-side synchronization: Analyzing client-side split network portion size to overall performancecited by 10
  14. 2025Dynamic Uncertainty-aware Multimodal Fusion for Outdoor Health Monitoringcited by 10
  15. 2022Actions at the Edge: Jointly Optimizing the Resources in Multi-access Edge Computingcited by 8
  16. 2025FSL-SAGE: Accelerating Federated Split Learning via Smashed Activation Gradient Estimationcited by 7
  17. 2024Faster Convergence on Heterogeneous Federated Edge Learning: An Adaptive Clustered Data Sharing Approachcited by 3
  18. 2024Parallel Split Learning with Global Samplingcited by 3
  19. 2025FedMobile: Enabling Knowledge Contribution-aware Multi-modal Federated Learning with Incomplete Modalitiescited by 3
  20. 2024Optimizing Personalized Federated Learning through Adaptive Layer-Wise Learningcited by 2
  21. 2025Data Heterogeneity and Forgotten Labels in Split Federated Learningcited by 2
  22. 2025Towards Collaborative Fairness in Federated Learning Under Imbalanced Covariate Shiftcited by 1

Built on by

Not cited yet.