Spatial-HTM: A MapReduce-Based System for Querying Spatial Data with the Hierarchical Triangular Mesh

摘要

Spatial data, in particular, spherical data, are essential for earth science, space science, astronomy, and any domains where observed objects are often located on the surface of the unit sphere. The volume challenge of such data and increasing importance for fine-granularity queries demand a framework that can handle big spherical data with advanced spherical index support. In this paper, we present Spatial-HTM, a MapReduce-based system with the Hierarchical Triangular Mesh (HTM) index support, and implement range querying of arbitrary convexes through the HTM preprocessing, operation, and storage modules of Spatial-HTM. The experiments show that Spatial-HTM outperforms Hadoop and SpatialHadoop with the Grid index in terms of both throughput and the number of returned points.