Regenerative LEO Satellite Architectures for Resilient 6G PPDR Communications

Kurzfassung

Large-scale natural disasters often damage terrestrial communication infrastructures, leading to major connectivity outages at a time when reliable communications are critical for public protection and disaster relief (PPDR) operations. While sixth-generation (6G) connectivity relies on low Earth orbit (LEO) non-terrestrial networks (NTNs), transparent satellite architectures schedule traffic at the ground segment, which restricts real-time adaptability under quickly changing disaster conditions. This article explores how resilient PPDR communications can be made possible by regenerative LEO NTN architectures with on-board resource management. A Stackelberg-based scheduling framework is adopted to efficiently manage limited satellite bandwidth while reinforcement learning (RL) enables dynamic policy adaptation in response to traffic surges. The proposed approach supports heterogeneous traffic with explicit prioritization of mission-critical emergency messages. The proposed framework is evaluated for regenerative satellite payloads with and without differentiated traffic handling using user satisfaction as the performance metric, together with latency, reliability evaluation and energy consumption considerations associated with on-board computation. The results indicate that priority-aware on-board resource management supports sustained service continuity for mission-critical users, while non-critical traffic experiences a gradual reduction in satisfaction under congestion.