Satellite Federated Fine-Tuning for Foundation Models in Space Computing Power Networks

2025-04-14

AI Summary

To address the challenges of limited on-board computing resources in low-Earth-orbit (LEO) satellites and the intermittent connectivity, short communication windows, and unstable links inherent in space networks, this paper proposes a space–ground collaborative federated fine-tuning framework enabling efficient in-orbit fine-tuning of large remote-sensing models. Methodologically, it introduces the first space–ground federated fine-tuning paradigm; achieves on-board lightweight computation via model modularization and distributed forward–backward propagation; and jointly optimizes computation and communication through three strategies: parallel intra-orbit communication, topology-aware space–ground scheduling, and low-latency inter-orbit transmission. Experimental results demonstrate a 33% reduction in training time, significantly enhancing the feasibility and timeliness of in-orbit fine-tuning for large models. This work establishes a scalable, distributed training pathway for onboard AI systems.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Computer Vision: Remote Sensing / Geospatial AI
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Advancements in artificial intelligence (AI) and low-earth orbit (LEO) satellites have promoted the application of large remote sensing foundation models for various downstream tasks. However, direct downloading of these models for fine-tuning on the ground is impeded by privacy concerns and limited bandwidth. Satellite federated learning (FL) offers a solution by enabling model fine-tuning directly on-board satellites and aggregating model updates without data downloading. Nevertheless, for large foundation models, the computational capacity of satellites is insufficient to support effective on-board fine-tuning in traditional satellite FL frameworks. To address these challenges, we propose a satellite-ground collaborative federated fine-tuning framework. The key of the framework lies in how to reasonably decompose and allocate model components to alleviate insufficient on-board computation capabilities. During fine-tuning, satellites exchange intermediate results with ground stations or other satellites for forward propagation and back propagation, which brings communication challenges due to the special communication topology of space transmission networks, such as intermittent satellite-ground communication, short duration of satellite-ground communication windows, and unstable inter-orbit inter-satellite links (ISLs). To reduce transmission delays, we further introduce tailored communication strategies that integrate both communication and computing resources. Specifically, we propose a parallel intra-orbit communication strategy, a topology-aware satellite-ground communication strategy, and a latency-minimalization inter-orbit communication strategy to reduce space communication costs. Simulation results demonstrate significant reductions in training time with improvements of approximately 33%.

Problem

• Addressing insufficient on-board computation for large foundation models in satellites
• Optimizing communication strategies for satellite-ground collaborative federated learning
• Reducing transmission delays in space computing power networks

Innovation

• Satellite-ground collaborative federated fine-tuning framework
• Parallel intra-orbit communication strategy reduces delays
• Topology-aware satellite-ground communication minimizes latency