基于HBase的高效交通数据云索引技术

Querying Research on Efficient Traffic Data Cloud-Indexing Technology Based on HBase

随着智能交通和物联网的发展,交通数据具有海量、多维、频繁更新等特征,传统数据库已无法满足查询效率需求。为提高查询效率,基于HBase提出一种支持高效更新和查询的交通数据索引框架。该框架采用三层索引结构,包括时间区间B+-树索引、子空间R-树索引、本地区域数据索引以支持数据更新,同时利用z-ordering技术进行数据分区以支持高效的多维查询。基于多样化数据集进行验证,实验证明所提方案在可拓展性和高效性方面均优于现有方案。

With the development of intelligent transportation and the Internet of Things, traffic data has the feature of massive, multi-dimensional and frequent updates, and traditional database has been unable to meet demand of query efficiency. To improve the query efficiency, we propose a traffic data indexing framework supporting efficient updating and querying based on HBase. The framework employs three-level index structure to update data quickly, which includes B+-tree Index for time intervals, R-tree index for subspaces, local index for data inside region. We also use the z-ordering technology which implements space partitioning to make the multi-dimension querying efficient. Based on the diverse data, we conduct our experiment and conclude that our scheme outperforms the existing scheme in terms of scalability and efficiency.