A Distributed Algorithm for Self-adaptive Routing in LEO Satellite Network

LEO satellite networks, represented by the successful Iridium System, are composed of multiple satellite nodes and inter-satellite links (ISL). Numerous routing algorithms have been designed to determine satisfying routes between data flow sources and destinations, within the constraints including delay, congestion control, and throughput and load balancing. This paper proposes a distributed network-state aware self-adaptive routing algorithm based on neighbor ISL status and node workload. Every satellite node is independently responsible for forwarding datagrams in its queue, with information about network status piggy-backed in the transmitted datagrams. Such information helps understand and predict the network workload status on each direction of the outgoing links, and is used for nearly-optimal selection of datagram outbound links to achieve load balancing and multi-path routing. Experiments are conducted on ns-2 simulation platform with a designed LEO walker, to implement and evaluate the effectiveness and efficiency of the proposed algorithm. The results show a significant improvement of more than 50% on the network workload balancing, with a few more hops in the selected multiple routing paths compared with the traditional Dijkstras shortest path algorithm.