Robust Software Defined Multicast in Broadband LEO Constellations

In broadband Low Earth Orbit (LEO) satellite networks, efficient video delivery is achieved through multicast techniques, which combine the bandwidth of both the satellite Access Networks and the Backbone Networks, resulting in significant bandwidth savings. However, previous research on multicast routing in LEO constellations has primarily focused on multicast in an ideal state, overlooking the challenges of obstacle avoidance posed by satellite failures or link congestion in real-world scenarios. In this paper, we introduce Software-Defined Multicast (SDM) to broadband LEO constellations and propose a robust SDM policy capable of effectively managing the dynamic nature of LEO constellation networks. To address the obstacle avoidance requirements, we develop an Obstacle-Avoiding Rectilinear Steiner Tree (OA-RST) algorithm and construct an intelligent multicast routing tree. Additionally, we present a resilient multicast management strategy to ensure seamless maintenance and management of multicast groups, even in the face of frequent satellite switching, dynamic membership updates, and resource contention among multiple groups. Simulation results demonstrate the effectiveness and efficiency of our proposed routing tree construction algorithm and management scheme in minimizing the total multicast tree length, conserving network bandwidth resources, and maintaining stable routing in the network.