Towards Resilient SDA: Graph Theory and Cooperative Control in Distributed Network Architectures

TL;DR

This work addresses the challenge of Space Domain Awareness in a congested, multi-orbit environment by proposing a distributed on-orbit architecture where heterogeneous satellites act as actuators within a shared SDA network. It introduces a graph-theoretic framework that combines Voronoi tessellations for spatial responsibility with Delaunay triangulations to define proximity-based communication paths, enabling scalable, dynamic coordination across LEO, MEO, HEO, and GEO. The study formalizes key graph metrics (hopcount, closeness, eccentricity, diameter, radius, degree, and coreness) and implements a simulation pipeline using Space-Track data and spaceflight tooling to evaluate topology evolution as actuator density changes. Findings show regime-dependent behavior: LEO networks become more efficient with added actuators, while higher orbits experience diminishing returns due to elongated inter-satellite links, though broader coverage persists; a hierarchical consensus model with approvers and verifiers is proposed to improve data governance and synchronization. Overall, the approach offers a practical pathway to resilient, decentralized SDA across growing constellations, with clear directions for validation and extension of role-based coordination and service metrics.

Abstract

Space Domain Awareness (SDA) involves the detection, tracking, and characterization of space objects through the fusion of data across the space environment. As SDA advances beyond localized or operator-specific capabilities, there is a growing reliance on in-domain space assets for real-time, distributed sensing and decision-making. This paper investigates the potential of on-orbit collaboration by enabling data sharing among heterogeneous satellites as actuators within a single orbital regime. Using graph-theoretic constructs, we define regions of spatial responsibility via Voronoi tessellations and model communication pathways between actuators using Delaunay triangulation. We apply this framework independently to Low Earth Orbit (LEO), Medium Earth Orbit (MEO), Highly Elliptical Orbit (HEO), and Geostationary Orbit (GEO), and analyze each to quantify structural properties relevant to efficient communication, cooperative control, and synchronization for SDA operations with the growth in deployments of space assets.

Figures (10)

• Figure 1: Distribution of active and inactive RSOs, and debris across orbital regimes (LEO, MEO, GEO, HEO and beyond) based on ASTRIAGraphESA2025SpaceEnvironment.
• Figure 2: Notional architecture for on-orbit distributed SDA with examples of current and future space surveillance systems such as GSSAP, NEOSSat, MW/MT, and Saphire, assuming roles of actuators with their respective capabilities for enhanced SDA across various orbital regimes.
• Figure 3: (a) Distribution of a set of actuators $A=30$ in a sphere. (b) $VD(A)$. (c) $DT(A)$.
• Figure 4: Voronoi diagrams illustrating spatial coverage of actuators in GEO and MEO.
• Figure 5: Performance metrics for LEO actuator network topology over time. (a) MST, (b) Degree variance, and (c) Average path length.