1: \begin{abstract}

2: 

3: We present \name, a stable space-ground network structure for low-earth orbit (LEO) satellite mega-constellations at scale. 

4: Due to the dynamic many-to-many space-ground mapping in high mobility, existing LEO mega-constellations with IP protocol stack suffer from 

5: %dynamic many-to-many mapping between space and ground, which results in 

6: frequent user IP address changes (every 133--510s per user) and network routing re-convergence (thus $\leq$20\% network usability). 

7: %To solve these fundamental stability issues in the highly dynamic environment, 

8: To provably stabilize the space-ground network under high mobility and many-to-many dynamics,

9: %\name aims to bridge the gap between the network design in the virtual cyberspace and the mega-constellations in the physical world.

10: \name adopts a recursive structure over the Rosette constellation, 

11: derives a hierarchical and time-invariant network address space from a new geographical coordinate system, 

12: and ensures efficient and stable routing via geographical-to-topological routing embedding without re-convergence.

13: %and eliminates routing re-convergence with local geographical-to-topological routing embedding over its stable topology.

14: %To this end, \name stabilizes the space-ground network topology, addressing and routing in such a highly dynamic environment. 

15: %\name adopts a recursive structure with provably stable topology and devises hierarchical and time-invariant addressing for space and ground. 

16: %It eliminates the routing re-convergence, and enables geographical routing for terrestrial users by converting it to routing over stable topology. 

17: Our hardware-in-the-loop, trace-driven emulations validate \name's stability, near-optimal routing ($\leq$1.4\% additional delays), and marginal overhead (<1\% CPU,  <2MB memory) for resource-constrained satellites.

18: 

19: \end{abstract}

20: