低轨大型星座分布式数据接收资源调度方法

许多监视与早期预警应用通过具备高时效性的对地观测数据实现。低轨大型星座观测数据接收任务调度的主要挑战是卫星数量大，中心化资源调度冲突消解面临困难。为此，提出一种新的分布式数据接收资源调度方法。首先，基于星间与星地链路融合构建卫星与地面设备之间的分布式传输链路，在卫星、设备、链路组成的无向网络图中，计算数据下传任务期间分布式链路的最短路径及其持续时长；其次，针对由每个设备的最短路径时间窗口合并构成的集合，利用Q学习方法求解数据接收完成率的优化问题。仿真测试结果表明，在典型的数据接收场景下，采用Q学习求解的数据接收完成率优于“先可见先跟踪”的基准资源调度方法，提升接近1%；与仅使用境内地面站的数据接收资源调度相比，数据接收完成率提升超过50%。

A Distributed Downlink Scheduling Approach to Large Low Earth Orbit Constellation

Many monitoring and early warning applications can be achieved with use of most up-o-ate Earth observation data.The main challenge with downlink scheduling for low Earth orbit (LEO) observation constellation is the large number of satellites.And it is hard to solve the centralized scheduling conflict.Therefore,a novel distributed downlink scheduling method is proposed.First,distributed downlinks between satellite and ground equipment merging inter-atellite link and satellite-round link are designed.The shortest path and its survival time of distributed downlink during the period of download mission are calculated based on an undirected graph consisted of satellite,equipment and downlink.Second,for the union of shortest path window from all ground equipment,Q-earning is introduced to solve the problem of reception rate optimization.The empirical results show that under typical data reception scenarios,the presented method using Q-earning achieves almost 1