Chingiz Akniyazov, University of Auckland; Roberto Armellin, University of Auckland; Aleksander Serebryanskiy, Fesenkov Astrophysical Institute
https://doi.org/10.64861/HNHX4777
Keywords: Cislunar SDA, Astrodynamics, Space Based Surveillance, Telescope Design, Optical Sensing, Space Domain Awareness, Sensors, CR3BP, Orbit Selection, Space Object Detection
Abstract:
The cislunar region is becoming an increasingly important operating environment for government, civil, and commercial missions, creating a growing need for physically defensible space-based surveillance architectures [1]. In this regime, observer placement cannot be treated as a purely astrodynamics problem, and payload selection cannot be treated as a purely optical design problem. Detection performance is governed by the coupled interaction of orbit geometry, line-of-sight access, illumination, target apparent motion, limiting magnitude, and instrument sampling over long-duration operations [1,2]. Prior studies have established the importance of cislunar visibility analysis [1], optical surveillance radiometry [2], and multi-objective telescope design for space situational awareness [3]. However, these elements are typically treated as separate trades, often applied to only a small number of candidate observer orbits, limited target populations, or reduced visibility criteria. As a result, a practical gap remains: there is no unified, validated framework that jointly selects observer orbit and telescope payload for cislunar surveillance as a single architecture-design problem.
This paper presents an observability-driven architecture for the coupled selection of observer orbit and telescope design for cislunar Space Domain Awareness. The proposed method begins by generating a synthetic cislunar target population distributed across multiple periodic orbit families in the Circular Restricted Three-Body Problem and propagating that population on a common stratified time grid [4]. A bounded three-dimensional search region is then populated with candidate observer anchor points, enabling the analysis to move beyond sparse manual orbit testing toward a scalable search over a large candidate space. For every anchor, target, and sampled time, the framework precomputes the key geometric observables governing optical detectability, including range, phase angle, Earth and Moon line-of-sight obstruction, eclipse conditions, and transverse angular velocity [1,2]. These quantities are used to construct a multi-threshold observability field in which each anchor is evaluated over a sweep of limiting-magnitude thresholds and object diameters. Rather than scoring performance using a single reference target size, the method forms a diameter-dependent completeness curve and applies a Jeffreys lower-bound statistical model to estimate conservative detection completeness under finite sampling [5]. The resulting completeness curve is collapsed into a scalar anchor score by integrating over logarithmic diameter, producing a threshold-dependent score terrain that captures both the physics of visibility and the statistics of confidence.
The anchor-level terrain is then elevated into an orbit-level decision process. Candidate observer orbit families are propagated on the same sampled time grid as the target population so that observer-target comparisons remain temporally consistent across the full pipeline. Each observer trajectory is projected onto the precomputed observability field through nearest-anchor mapping with a geometric support gate, producing a time history of along-track completeness as a function of target diameter. These histories are reduced into orbit-level summary curves and area-based performance metrics that quantify how consistently an observer traverses high-value surveillance regions over time. Winner selection is performed independently for each limiting-magnitude threshold using a distributional rule designed to avoid fragile optima: instead of selecting the single numerically largest score directly, the framework first identifies the highest occupied performance class and then resolves ties using a tracking-difficulty proxy based on the spread of maximum transverse angular rate, favoring observer orbits that are both high performing and operationally trackable [6]. This produces a more stable and mission-relevant ranking than a strict pointwise maximum, particularly when several candidate orbits occupy a nearly equivalent top-performance regime.
To ensure that the reduced-order ranking remains faithful to the underlying physics, the selected orbit is subjected to a dedicated truth-validation stage. In this stage, direct radiometric and geometric calculations are performed against the full target population to compute pooled conservative completeness and time-averaged detectable fraction across diameter, enabling an apples-to-apples comparison between the anchor-based surrogate and the full truth model [2]. The validated operating environment is then used to drive separate telescope design studies for survey and tracking operations over aperture, f-number, pixel size, and exposure time [2,3,6]. Survey-oriented payloads are assessed primarily through detection completeness and achievable sky-coverage rate, while tracking-oriented payloads are considered separately according to their cadence and image-sampling performance across the relevant apparent-motion environment. The result is a quantitatively validated, end-to-end architecture that unifies astrodynamics, cislunar SDA, and optical sensor design into a single decision framework for generating traceable orbit-telescope recommendations in the cislunar regime.
Date of Conference: September 15-18, 2026
Track: Cislunar SDA