一种支持高效并行处理的矢量数据索引方法

发布时间：2018-06-15 07:43

本文选题：spark + hilbert　；参考：《计算机工程与应用》2017年11期

【摘要】：分析了HBase的存储模型和Spark的并行处理机制,提出一种矢量空间数据的分布式存储、索引和并行区域查询方法。设计了基于空间对象中心点的行键存储方案,将中心点的Hilbert编码与经纬度小数位结合实现行键的唯一性,保证地理位置接近的要素在表中存储在相邻的行。实现了基于Spark的空间索引并行构建和区域查询方法,借助空间对象中心点的Hilbert编码快速构建索引,通过多边形区域的最小外接矩形过滤查询结果。实验结果表明,索引并行构建可靠性好速度快,区域查询并行处理算法可行且效率高。
[Abstract]:This paper analyzes the storage model of HBase and the parallel processing mechanism of Spark, and proposes a distributed storage, index and parallel region query method for vector spatial data. A row key storage scheme based on the center point of the spatial object is designed. The Hilbert coding of the center point is combined with the latitude and longitude decimal places to realize the uniqueness of the line key, which ensures that the elements close to the geographical position are stored in the adjacent rows in the table. The spatial index parallel construction and the region query method based on Spark are implemented. The index is constructed quickly by Hilbert coding of the center point of the spatial object, and the query results are filtered by the minimum external rectangle of the polygon region. The experimental results show that the parallel construction of index is reliable and fast, and the parallel processing algorithm of region query is feasible and efficient.
【作者单位】：平顶山学院软件学院;武汉大学软件工程国家重点实验室;桂林理工大学南宁分校;平顶山学院资源与环境科学学院;
【基金】：国家自然科学基金青年科学基金(No.41201405) 河南省科技厅科技攻关项目(No.152102310202)
【分类号】：TP333;TP391.3

【相似文献】