A Density-aware SDN Controller Placement in LEO Satellite Networks
Syed M. A. H. Bukhari, Muhammad Asif, Umar Mahmood, Wang‐Cheol Song · 2025
The rapid deployment of Low Earth Orbit (LEO) satellite constellations necessitates efficient network management solutions due to their dynamic topologies and high node mobility. Software-Defined Networking (SDN) offers a promising architecture for centralized control and programmability in such environments. However, optimal controller placement remains a key challenge due to the time-varying nature of satellite positions and traffic demands. In this paper, we propose a density-aware SDN controller placement strategy that considers local satellite density, isolation distance, and neighborhood compactness to select controllers that are both traffic-aware and spatially welldistributed. The proposed approach employs a greedy selection mechanism followed by refinement to minimize worst-case switch to controller delays. We evaluate the proposed method using the Iridium-NEXT constellation and compare its performance with baselines, including Standard K-Means, K-Means++, and Spectral Clustering. Experimental results show that the proposed method significantly outperforms the baselines in terms of the worst-case switch to controller delay, achieving up to a $93 \%$ improvement.