Quadrant-Based MBR-Tree Indexing Technique for Range Query Over HBase

초록

HBase is one of the most popular NoSQL database systems. Because it operates on a distributed file system and supports a flexible schema, it is suitable for dealing with large volumes of semi-structured data. However, HBase only provides an index built on one dimensional rowkeys of data, which is unsuitable for the effective processing of multidimensional spatial data. In this paper, we propose a hierarchical index structure called a Q-MBR (quadrant based minimum bounding rectangle) tree for effective spatial query processing in HBase. We construct a Q-MBR tree by grouping spatial objects hierarchically through Q-MBRs. We also propose a range query processing algorithm based on the Q-MBR tree. Our proposed range query processing algorithm reduces the number of false positives significantly. An experimental analysis shows that our method performs considerably better than the existing methods.

키워드