Instability of Self-Driving Satellite Mega-Constellation: From Theory to Practical Impacts on Network Lifetime and Capacity
Yimei Chen, Yuanjie Li, Hewu Li, Lixin Liu, Li Ouyang, Jiabo Yang, Junyi Li, Jianping Wu, Qian Wu, Jun Liu, Zeqi Lai
TL;DR
This work identifies an endogenous instability in self-driving LEO mega-constellations where local pairwise collision-avoidance policies trigger cascaded internal maneuvers, risking network lifetime and capacity. It develops a Lyapunov-based stability framework, derives a stability condition $\alpha_2^2 - \alpha_3^2 - 2\alpha_1 \ge 0$, and shows how unstable policies cause explosive maneuver growth via the transfer function $H(\lambda)$. Empirical validation using Starlink space situational awareness data reveals a high incidence of cascaded maneuvers (e.g., 81.4% of Starlink-on-Starlink maneuvers) and substantial lifetime impacts, with cascades consuming the majority of maneuver budgets. As a practical remedy, bilateral maneuver control adds a backward coupling to the existing forward policy, leading to stable dynamics and enabling simultaneous improvements in network lifetime and capacity; trace-driven evaluation reports up to an $8\times$ lifetime extension with no loss in capacity and about an $80\%$ reduction in maneuver amplification.
Abstract
Low Earth Orbit (LEO) satellite mega-constellations aim to enable high-speed Internet for numerous users anywhere on Earth. To safeguard their network infrastructure in congested outer space, they perform automatic orbital maneuvers to avoid collisions with external debris and satellites. However, our control-theoretic analysis and empirical validation using Starlink's space situational awareness datasets discover that, these safety-oriented maneuvers themselves can threaten safety and networking via cascaded collision avoidance inside the mega-constellation. This domino effect forces a dilemma between long-term LEO network lifetime and short-term LEO network capacity. Its root cause is that, the decades-old local pairwise maneuver paradigm for standalone satellites is inherently unstable if scaled out to recent mega-constellation networks. We thus propose an alternative bilateral maneuver control that stabilizes self-driving mega-constellations for concurrent network lifetime and capacity boosts. Our operational trace-driven emulation shows a 8$\times$ network lifetime extension in Starlink without limiting its network capacity.