A Graph Koopman Autoencoder enables scalable SDN control for LEO mega-constellations by compressing topology and forecasting dynamics, showing 42.8% better spatial compression and 10.81% better temporal forecasting than baselines in Starlink simulations.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Toward Scalable SDN for LEO Mega-Constellations: A Graph Learning Approach
A Graph Koopman Autoencoder enables scalable SDN control for LEO mega-constellations by compressing topology and forecasting dynamics, showing 42.8% better spatial compression and 10.81% better temporal forecasting than baselines in Starlink simulations.