An Adaptive Routing Algorithm for Inter-Satellite Networks Based on the Combination of Multipath Transmission and Q-Learning

In a satellite network, the inter-satellite link can facilitate information transmission and exchange between satellites, packet routing of is key development direction communication systems. Aiming at complex topology dynamic change in LEO networks, traditional single shortest path algorithm no longer meet optimal requirement. Therefore, this paper proposes multi-path based on an improved breadth-first search. First, according to network information, search used obtain all front hop node destination node. Second, paths are obtained by backtracking through Finally, bandwidth capacity traffic nodes determines from multiple paths. However, due high dynamics low-orbit changes rapidly, global often not available. At time, order enhance adaptability algorithm, reinforcement learning. Verified simulation experiments, proposed improve throughput reduce time delay, loss rate.