LEO-Split: A Semi-Supervised Split Learning Framework over LEO Satellite Networks
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913633530806272 |
|---|---|
| author | Lin, Zheng Zhang, Yuxin Chen, Zhe Fang, Zihan Wu, Cong Chen, Xianhao Gao, Yue Luo, Jun |
| author_facet | Lin, Zheng Zhang, Yuxin Chen, Zhe Fang, Zihan Wu, Cong Chen, Xianhao Gao, Yue Luo, Jun |
| contents | Recently, the increasing deployment of LEO satellite systems has enabled various space analytics (e.g., crop and climate monitoring), which heavily relies on the advancements in deep learning (DL). However, the intermittent connectivity between LEO satellites and ground station (GS) significantly hinders the timely transmission of raw data to GS for centralized learning, while the scaled-up DL models hamper distributed learning on resource-constrained LEO satellites. Though split learning (SL) can be a potential solution to these problems by partitioning a model and offloading primary training workload to GS, the labor-intensive labeling process remains an obstacle, with intermittent connectivity and data heterogeneity being other challenges. In this paper, we propose LEO-Split, a semi-supervised (SS) SL design tailored for satellite networks to combat these challenges. Leveraging SS learning to handle (labeled) data scarcity, we construct an auxiliary model to tackle the training failure of the satellite-GS non-contact time. Moreover, we propose a pseudo-labeling algorithm to rectify data imbalances across satellites. Lastly, an adaptive activation interpolation scheme is devised to prevent the overfitting of server-side sub-model training at GS. Extensive experiments with real-world LEO satellite traces (e.g., Starlink) demonstrate that our LEO-Split framework achieves superior performance compared to state-ofthe-art benchmarks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_01293 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LEO-Split: A Semi-Supervised Split Learning Framework over LEO Satellite Networks Lin, Zheng Zhang, Yuxin Chen, Zhe Fang, Zihan Wu, Cong Chen, Xianhao Gao, Yue Luo, Jun Machine Learning Artificial Intelligence Distributed, Parallel, and Cluster Computing Networking and Internet Architecture Recently, the increasing deployment of LEO satellite systems has enabled various space analytics (e.g., crop and climate monitoring), which heavily relies on the advancements in deep learning (DL). However, the intermittent connectivity between LEO satellites and ground station (GS) significantly hinders the timely transmission of raw data to GS for centralized learning, while the scaled-up DL models hamper distributed learning on resource-constrained LEO satellites. Though split learning (SL) can be a potential solution to these problems by partitioning a model and offloading primary training workload to GS, the labor-intensive labeling process remains an obstacle, with intermittent connectivity and data heterogeneity being other challenges. In this paper, we propose LEO-Split, a semi-supervised (SS) SL design tailored for satellite networks to combat these challenges. Leveraging SS learning to handle (labeled) data scarcity, we construct an auxiliary model to tackle the training failure of the satellite-GS non-contact time. Moreover, we propose a pseudo-labeling algorithm to rectify data imbalances across satellites. Lastly, an adaptive activation interpolation scheme is devised to prevent the overfitting of server-side sub-model training at GS. Extensive experiments with real-world LEO satellite traces (e.g., Starlink) demonstrate that our LEO-Split framework achieves superior performance compared to state-ofthe-art benchmarks. |
| title | LEO-Split: A Semi-Supervised Split Learning Framework over LEO Satellite Networks |
| topic | Machine Learning Artificial Intelligence Distributed, Parallel, and Cluster Computing Networking and Internet Architecture |
| url | https://arxiv.org/abs/2501.01293 |