基于SDN的高性能QoS保障低轨道卫星星间路由算法

低轨道卫星通信系统具有全球覆盖性、移动性、可扩展性等优势，在提供全球互联网服务、灾难应急处理等方面发挥重要作用，但由于星上有限的存储和计算资源，传统路由算法不适用于低轨道卫星通信网络。结合软件定义网络架构，提出一种支持服务质量（QoS）的高性能低轨道卫星星间路由算法。根据剩余链路持续时间定义星间链路生存时间，确定每条星间链路的稳定度，缓解由于链路切换导致的业务路径重构问题。基于高轨道卫星得到的星间链路的流量状态，定义链路负载矩阵，给出星间链路负载度函数，并利用标签交换路径集合获得每条路径的负载度，避免节点拥塞，实现网络负载均衡。针对不同要求的业务服务类型定义权重因子矩阵，通过调整因子来减小瓶颈节点对路由算法的影响，满足多用户的QoS要求。仿真结果表明，在不同的QoS要求下，该算法在业务时延、系统吞吐量、网络负载均衡等方面均具有明显优势，且算法复杂度低，大幅节省了有限的星上存储与计算资源。

SDN-Based High-Performance and QoS Guaranteed Inter-Satellite Routing Algorithm for Low-Earth Orbit Satellites

The Low-Earth Orbit(LEO) satellite communication system has the advantages of good global coverage, mobility, and scalability.It has played a significant role in providing global Internet services and disaster emergency response.However, owing to the limited storage and computing resources of satellites, traditional routing algorithms are unsuitable for LEO communication networks.Therefore, combined with Software-Defined Network(SDN) architecture, this study proposes a high-performance and Quality of Service(QoS) guaranteed inter-satellite routing algorithm for LEO satellites.We define the survival time of the Inter-Satellite Link(ISL) according to the remaining link duration and obtain the stability degree of each ISL to alleviate the problem of service path reconstruction caused by link switching. Based on the traffic status of the ISLs obtained by Geostationary-Earth Orbit(GEO) satellites, we define the link load matrix to provide the ISL load degree function, and use the Label Switching Path(LSP) set to obtain the load degree of each path to avoid node congestion and realize network load balancing.Finally, we define a weighting factor matrix for different required business service types and reduce the influence of the bottleneck node on the routing algorithm by adjusting the factor and guaranteeing the QoS requirements of multiple users.The simulation results show that under different QoS requirements, the algorithm has obvious advantages in terms of business delay, system throughput, and network load balancing, and the low algorithm complexity significantly reduces the limited on-board storage and computing resources.