Storing and Querying Semi-structured Spatio-Temporal Data in HBase

With the development of remote sensing, positioning and other technology, a large amount of spatio-temporal data require effective management. In the current research status, a lot of works have focused on how to effectively use HBase to store and quickly find structured spatio-temporal data. However, some spatio-temporal data exists in the semi-structured documents, such as metadata that describes the remote sensing products, under such context, the query is changed to spatio-temporal query + semi-structured query (XPath), which is less studies in previous works. In this paper, we focus on how to efficiently and economically achieve semi-structured spatio-temporal data storage and query in HBase. Firstly, the formal description of the problem is presented. Secondly, we propose HSSST storage model using a semi-structured approach TwigStack. On this basis, semi-structured spatio-temporal range query and kNN queries are carried out. Experiments are conducted on real dataset, comparing with MongoDB which need higher hardware configuration, the results show that in moderate configuration of machines, the performance of semi-structured spatio-temporal query algorithms are superior to MongoDB, thus it has advantage in real application.