Performance Analysis of Algorithms for Coordination of Earth Observation by CubeSat Constellations

Abstract

Two algorithms that coordinate activities across a resource-constrained, Earth-observing CubeSat constellation are presented. The Resource-Aware SmallSat Planner algorithm performs online planning of activities for a satellite while keeping the satellite’s resources within constraints. The Limited Communication Constellation Coordinator algorithm performs coordination of observations across the constellation to reduce average revisit times for a set of targets. The algorithms are simulated for 24 h with an 18 satellite Low Earth Orbit constellation. Three orbital geometries are examined, with different configurations of intersatellite and satellite-to-ground communications links to share planning information. Results indicate that coordination through Resource-Aware SmallSat Planner/Limited Communication Constellation Coordinator and a background communications constellation improves observation performance, with sensor-averaged revisit times of 197, 203, and 225 min (over three orbital geometries) versus 204, 211, and 240 min for a baseline random sensor selection method. The results also reveal that the constellations studied perform poorly at sharing planning information via only downlinks and crosslinks and point to the need for a method of calculating the information sharing utility of such links in a coordinated constellation. These findings are relevant for Federated Satellite Systems because they provide guidance on next steps toward integrating large networks of heterogeneous small satellites into large-scale, coordinated, Earth-observing systems.