Concurrent Optimization of Satellite Phasing and Tasking for Cislunar Space Situational Awareness

TL;DR

This work addresses the concurrent optimization of satellite phasing and tasking for cislunar space situational awareness using CR3BP dynamics and EKF-based estimation. It introduces two information-theoretic objectives, Max and MaxMin, and a two-level optimization where the upper-level phasing vector $\boldsymbol{x}$ informs the lower-level tasking decisions $u_{ijk}$. The authors show gradient-based methods effectively optimize the upper level, validate a greedy approach for the Max objective, and reveal trade-offs: MaxMin improves target diversification at the cost of increased computation. Taken together, the results offer practical guidance for designing scalable, space-based SSA constellations in the Earth-Moon system, balancing coverage, fairness, and computational tractability.

Abstract

Recently, renewed interest in cislunar space spurred by private and public organizations has driven research for future infrastructure in the region. As Earth-Moon traffic increases amidst a growing space economy, monitoring architectures supporting this traffic must also develop. These are likely to be realized as constellations of patrol satellites surveying traffic between the Earth and the Moon. This work investigates the concurrent optimization of patrol satellite phasing and tasking to provide information-maximal coverage of traffic in periodic orbits.

Figures (8)

• Figure 1: The logarithm of the objective plotted against the phase of the first observer (left) and the second observer (right). The phase is parameterized as a fraction of the period of the observer.
• Figure 2: Contours of the objective in the entire search space (left) and near the optimizer (right)
• Figure 3: The logarithm of the objective plotted against the phase of an observer. The phase of the second observer is held constant. Although the scale of the information gain changes, the location of the optimal phase for observer 1 remains the same.
• Figure 4: Agent/Target orbits (left) and their projections on the xy plane (top right) and xz plane (bottom right). Orbits were chosen carefully to ensure that no intersection occurs between an agent and target orbit.
• Figure 5: Agent/Target Orbits (left) chosen intentionally to intersect. Objective dependence on the observer phase (right). Peaks correspond to square markers on the orbit plot (left), where an agent and target intersect.