LLM agent improves routing with dynamic quality service in LEO satellites
STR-Agent: An LLM-Driven Agent for QoS-Aware Routing in LEO Satellite Networks
Networking and Internet ArchitectureArtificial Intelligence
Summary
LEO satellite networks are tricky because their connections change all the time, and different users need different service types. The authors made STR-Agent, a smart system that understands natural language requests and turns them into routing instructions customized for network conditions. It learns from past results and adjusts on the fly to keep the network running smoothly. In simulated tests, STR-Agent lowered delays and better understood user needs compared to older methods.
What this means in practice
- •For network operators: Improve satellite network routing to meet varied user service needs dynamically using natural language inputs.
- •For satellite communications engineers: Design routing protocols for LEO constellations that adjust in real time based on network congestion and service priorities.
- •For cloud service providers: Use LLM-driven agents to translate client service requests into optimized satellite network routing configurations.$Commercial implications: Enables marketable adaptive QoS routing solutions that can automatically tailor satellite connectivity for diverse clients.
Authors
Bowen Lu, Mugen Peng, Yaohua Sun, Hongyu Wang, Kerui Guo, Wenjia Xu
Abstract
LEO satellite networks feature dynamic topologies, time-varying links, and diverse service requirements, which make conventional routing schemes difficult to support fine-grained quality-of-service (QoS) provisioning. Existing studies mainly optimize routing over network states with predefined objectives, but rarely address the practical challenge of translating unstructured natural-language service requests into adaptive routing decisions. To bridge this gap, we propose STR-Agent, an LLM-driven framework for QoS-aware routing in LEO satellite networks. The key innovation of STR-Agent lies in unifying intent perception, tool-based execution, experience accumulation, and reflection-based policy adaptation within a single agent architecture. Specifically, the Perception Module converts natural-language requests into structured routing semantics, while the Reflection Module dynamically adjusts the service-to-routing-policy mapping according to real-time congestion conditions and historical routing outcomes, rather than relying on a fixed routing objective. In addition, we develop a specialized perception model, and construct a domain-specific supervised fine-tuning dataset for LEO service understanding. Simulation results in a Walker-Delta constellation show that STR-Agent significantly outperforms conventional baselines: it reduces end-to-end delay by up to 60% compared with DQ-Dijkstra, improves average intent-understanding accuracy from 45.4% to 92.45% after supervised fine-tuning, and the Reflection Module further reduces the delay by 120 ms at 600 Mbps. These results demonstrate the potential of LLM-driven agent architectures to enable service-aware and adaptive QoS routing in future LEO satellite networks.