Research on the Application of Software-Defined Networking in Low Earth Orbit Satellite Remote Sensing Communications

This paper presents a comprehensive solution for integrating Satellite de Sensoriamento Remoto (SSR) into LEO satellite networks. It uses the Task-Edge Graph (TEG) for modeling task flows and resources, and leverages Virtual Network Function (VNF) orchestration from Network Function Virtualization (NFV). To address static unicast routing challenges like uneven traffic distribution and network congestion, two innovative routing algorithms are introduced.The Node Disjoint-based Multipath (NDM) algorithm designs primary and backup paths based on the shortest propagation distance, allowing adaptive route switching. The Minimum Dynamic Cost (MDC) algorithm considers queuing, node processing delay, and path propagation delay to calculate dynamic costs, reducing resource wastage during route discovery. The paper provides a system model, introduces algorithms to address the challenges, and conducts comprehensive performance analysis through simulations. This research enhances SSR communication efficiency and reliability in LEO satellite networks, with potential applications in various scenarios.