Hongseok Kim, University of Texas Austin; Maruthi Akella, University of Texas Austin
https://doi.org/10.64861/WQXB8795
Keywords: , LEO Mega-Constellations, Ground Sensor Scheduling, Mixed Integer Linear Programming, Distributed Optimization
Abstract:
The rapid growth of Low Earth Orbit (LEO) constellations has introduced increasingly complex scheduling and resource-allocation challenges across Earth observation, satellite communications, and Space Situational Awareness (SSA) missions. Although orbital motion can be accurately propagated over a planning horizon, satellite-ground access opportunities remain intermittent, and both spacecraft and ground assets are subject to operational limitations such as payload duty cycle, sensing cadence, antenna occupancy, and recovery time.
This paper presents a generalized mixed-integer optimization framework for selecting feasible satellite-ground contacts while minimizing the temporal service gaps experienced by mission targets. The formulation represents local operation-cadence constraints and target-side revisit-time objectives using fixed selection and difference matrices, allowing the same scheduling structure to be applied in both satellite-to-ground (SAT-to-GS) and ground-to-satellite (GS-to-SAT) directions. Contact maximization, square-sum revisit-time minimization, and maximum-revisit-time minimization objectives are considered, together with sequential block-coordinate descent (BCD) and heuristic consensus alternating direction method of multipliers (ADMM) solution strategies.
Mission heterogeneity is incorporated by adjusting resource-dependent cadence parameters, target-dependent weighting gains, and the number of parallel payload or ground-resource channels. The proposed framework is demonstrated through Earth-observation image acquisition, 6G satellite gateway scheduling with multi-beam resources, distributed ground-asset scheduling for SSA, and end-to-end Earth-observation image-data download analysis.
Numerical results illustrate the trade-offs among resource availability, operational cadence, target priority, and revisit-time performance, demonstrating the applicability of the proposed framework to a broad range of LEO constellation operations.
Date of Conference: September 15-18, 2026
Track: Space Domain Awareness