Relationship Between Robustness Metrics and Attack Tolerance in Satellite Communication Constellations

Abstract

The rapid deployment of massive low Earth orbit satellite communication constellations has revolutionized global internet connectivity, providing high bandwidth and low latency coverage to underserved regions. However, the increasing reliance on these space based networks introduces critical vulnerabilities to both random failures and intentional cyber or kinetic attacks. This paper investigates the complex relationship between static network robustness metrics and dynamic attack tolerance within simulated satellite communication constellations. By modeling large scale space networks as complex graphs, this study evaluates how traditional topological metrics, including average node degree, betweenness centrality, and algebraic connectivity, predict the network capacity to withstand various attack vectors. Utilizing an advanced discrete event simulation environment, we subjected multiple constellation topologies to both random node failures and targeted centrality based attacks. The structural and functional degradation of the network was measured continuously to determine the correlation between initial robustness metrics and the threshold of network fragmentation. The simulation evidence reveals that while high average degree improves tolerance against random failures, algebraic connectivity serves as a vastly superior predictor of resilience against targeted attacks aimed at critical routing hubs. The findings provide critical insights for constellation designers, suggesting that optimizing for specific spectral graph properties can significantly enhance the survivability of next generation space networks without requiring proportional increases in physical redundancy or launch costs.