Spatiotemporal Indexing and Query Application on Cassandra for Large-Scale Trajectory Data

Abstract

The fast-growing urban traffic produces a huge volume of traffic data which becomes a problem when storing, querying and analyzing large-scale data in traditional database technologies. As a viable alternative, NoSQL databases offer high scalability and high input/output throughput features which potentially accommodate the needs. We propose a spatiotemporal indexing method which can be applied to the Cassandra database for efficient indexing and storing of large-scale spatiotemporal datasets aiming to provide an effective framework for processing, querying, and analyzing large amounts of data. We improve the Z-order curve by adding temporal information as the third dimension and then encode the resultant's values with the Geohash standard. The resulting spatiotemporal hash code is applied to the Cassandra database. A corresponding indexing structure is designed so that the Cassandra database will physically store the spatiotemporally neighbouring data points close to each other. Such a method aims to increase query efficiency in large-scale spatiotemporal data querying and analyzing applications. The dataset acquired from the real applications is used to conduct the computational experiments to validate the efficiency of the proposed approach and benchmark against some existing database technologies. The computational experiments reveal the superiority of the proposed method compared to the ordinary methodologies, The efficiency of the proposed methodology is validated further by applying it to query the vehicle trajectories gathering the real-time air quality data. The query experiments on two benchmarks illustrate that spatiotemporal indexing sig-nificantly improves the query performance over the ordinary methodology. In addition, spatiotemporal indexing can provide a stable performance as dataset size increases.