Flow Granularity Multi-path Transmission Optimization Design for Satellite Networks.

The natural mesh network topology of satellite networks makes multi-path transmission prevalent due to its ability to provide bandwidth aggregation and backup using redundant paths. While as one of the significant benefits of multi-path transmission, high reliability requires extensive signaling to achieve excellent performance. So designing an analysis and treatment method to tackle the network overhead and reliability balance remains a major challenge. In addition, massive and burst services in satellite networks will have different demands, which require the network to execute fine-grained path scheduling and management of the multi-path transmission. In this paper, we carefully model the multi-path reliability and network overhead factors under the SDN-based integrated satellite-terrestrial network architecture to characterize the network state. Second, an adaptive multi-path selection scheme is designed to handle the number of active paths, which considers different Quality of Service (QoS) requirements under the limited resources of satellite networks. Then, we formulate the problem as Non-Linear Binary Programming (NLBP) and develop a practical Particle Swarm Optimization (PSO)-based algorithm to solve it. Simulation results show that the proposed scheme can improve satellite networks' throughput and resource utilization.