A Resilient and Efficient Low-Latency Hybrid Routing Algorithm for LEO Constellations

Low Earth Orbit (LEO) satellite constellations are emerging as a key enabler of next-generation global communications, but achieving low-latency and reliable routing across inter-satellite links (ISLs) remains a major challenge. This paper proposes a hybrid routing strategy, termed Hybrid DCLT-A*, which combines a Delay-Constrained Longest Lifetime (DCLT) metric for satellite access point selection with the heuristic efficiency of the A* algorithm for inter-satellite pathfinding. Using simulations over a realistic Iridium-like constellation (66-satellite reference geometry), the algorithm demonstrates near-optimal delay performance under dynamic ISLs while improving path stability, achieving a mean propagation delay of 49.17 ms under dynamic ISLs, compared with 48.77 ms for Dijkstra and A*. Under fixed ISLs, the three methods have the same reported peak, mean, and RMS propagation delays of 57.95, 50.97, and 51.11 ms, respectively. Under fixed ISLs, it matches state-of-the-art schemes and exhibits superior robustness to satellite and link failures. The framework is extended to include user priorities, payload and ISL health, link budget feasibility, bandwidth, ISL bitrate, and transmit power constraints, enabling adaptive, resource-aware routing under degraded network conditions, including cases with 100 users and up to 20 gateways. They do not establish reduced access handovers, superior comparative resilience, or large-scale computational performance.