# Split Federated Learning Architectures for High-Accuracy and Low-Delay Model Training

arXiv: https://arxiv.org/abs/2603.08687 · HTML: https://mlsanity.com/paper/2603.08687/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 42 references that are in the map.
- First posted on arXiv 9 Mar 2026.

## Builds on (most cited first, 21 of 21)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Federated Learning with Non-IID Data](https://mlsanity.com/paper/1806.00582/) (2018)
- [Split learning for health: Distributed deep learning without sharing raw patient data](https://mlsanity.com/paper/1812.00564/) (2018)
- [SplitFed: When Federated Learning Meets Split Learning](https://mlsanity.com/paper/2004.12088/) (2020)
- [Heterogeneous Federated Learning: State-of-the-art and Research Challenges](https://mlsanity.com/paper/2307.10616/) (2023)
- [Training Neural Networks with Local Error Signals](https://mlsanity.com/paper/1901.06656/) (2019)
- [Efficient Parallel Split Learning over Resource-constrained Wireless Edge Networks](https://mlsanity.com/paper/2303.15991/) (2023)
- [Asynchronous Federated Learning on Heterogeneous Devices: A Survey](https://mlsanity.com/paper/2109.04269/) (2021)
- [TernausNet: U-Net with VGG11 Encoder Pre-Trained on ImageNet for Image Segmentation](https://mlsanity.com/paper/1801.05746/) (2018)
- [The History Began from AlexNet: A Comprehensive Survey on Deep Learning Approaches](https://mlsanity.com/paper/1803.01164/) (2018)
- [Split Learning over Wireless Networks: Parallel Design and Resource Management](https://mlsanity.com/paper/2204.08119/) (2022)
- [AdaptSFL: Adaptive Split Federated Learning in Resource-constrained Edge Networks](https://mlsanity.com/paper/2403.13101/) (2024)
- [FedAvg with Fine Tuning: Local Updates Lead to Representation Learning](https://mlsanity.com/paper/2205.13692/) (2022)
- [Hierarchical Split Federated Learning: Convergence Analysis and System Optimization](https://mlsanity.com/paper/2412.07197/) (2024)
- [A Bargaining Game for Personalized, Energy Efficient Split Learning over Wireless Networks](https://mlsanity.com/paper/2212.06107/) (2022)
- [Workflow Optimization for Parallel Split Learning](https://mlsanity.com/paper/2402.10092/) (2024)
- [FedSplitX: Federated Split Learning for Computationally-Constrained Heterogeneous Clients](https://mlsanity.com/paper/2310.14579/) (2023)
- [MP-SL: Multihop Parallel Split Learning](https://mlsanity.com/paper/2402.00208/) (2024)
- [Collaborative Split Federated Learning with Parallel Training and Aggregation](https://mlsanity.com/paper/2504.15724/) (2025)

## Abstract (arXiv)

Can we find a network architecture for ML model training so as to optimize training loss (and thus, accuracy) in Split Federated Learning (SFL)? And can this architecture also reduce training delay and communication overhead? While accuracy is not influenced by how we split the model in ordinary, state-of-the-art SFL, in this work we answer the questions above in the affirmative. Recent Hierarchical SFL (HSFL) architectures adopt a three-tier training structure consisting of clients, (local) aggregators, and a central server. In this architecture, the model is partitioned at two partitioning layers into three sub-models, which are executed across the three tiers. Despite their merits, HSFL architectures overlook the impact of the partitioning layers and client-to-aggregator assignments on accuracy, delay, and overhead. This work explicitly captures the impact of the partitioning layers and client-to-aggregator assignments on accuracy, delay and overhead by formulating a joint optimization problem. We prove that the problem is NP-hard and propose the first accuracy-aware heuristic algorithm that explicitly accounts for model accuracy, while remaining delay-efficient. Simulation results on public datasets show that our approach can improve accuracy by 3%, while reducing delay by 20% and overhead by 50%, compared to state-of-the-art SFL and HSFL schemes.