Topology Awareness for LEO Satellite Networks with Limited Ground Station Resources

Abstract

Accurate topology awareness of non-cooperative low Earth orbit (LEO) satellite constellations faces two major challenges: limited ground observation resources and the unavailability of internal node cooperation. To address these issues, this paper proposes a two-stage framework that jointly identifies physical topology indicated by satellite orbital parameters and logical topology indicated by tree-structured forwarding paths under territorial ground station constraints. The first stage employs a geometric orbit determination algorithm using multi-station synchronous ranging data. Ground stations are deployed following an “east-west elongated, multi-point distributed” strategy to maximize observation baselines while ensuring coverage reliability. Satellite positions are solved via Levenberg-Marquardt minimization, and Keplerian elements are fitted under circular orbit assumptions. The second stage introduces the Layer-based Network Inference Heuristic Algorithm (LNIHA), which combines Time-to-Live (TTL) probing for hop count acquisition and Sandwich Probe techniques for shared path length inference from a single source node without internal cooperation. Simulation tests are conducted on the proposed algorithm. The simulation results demonstrate that joint physical and logical topology awareness is feasible under realistic operational constraints, providing a foundation for non-cooperative LEO constellation monitoring and spectrum management.