多用户全球三维仿真的数据组织与调度研究

Study of Data Organization and Scheduling Method in Multi-user Accessed Global 3D Visualization Application

以多用户全球三维仿真为应用背景,研究P2P环境下海量数据的多点并行分发技术。通过依据地理空间的影像分块、编码协议和Tracker服务器所维护的资源拥有节点四叉树索引列表,实现P2P网络中的快速资源定位;利用基于数据预测、缓冲思想的优先下载策略和用户节点间的直接数据交换,提高数据分发效率。最后,通过对现有的多用户三维地形可视化系统进行改造试验,证明该方法的有效性。

The massive remote sensing data＇s parallel dispensing method in multi users accessed global 3D visualization application was studied based on P2P technology.By using image pitching according to geographic space,encoding protocol and Quad-Tree index,data resource＇s quick searching was realized,and by making use of prior downloading strategy based on data forecast,cache and direct data exchanging between peers,data dispensing efficiency was improved.At last,a system experiment was made to prove this method.