MapReduce架构下的大规模轨迹数据压缩策略

Compression Strategy of Large Scale Trajectory Data Based on MapReduce Architecture

车辆GPS轨迹数据中蕴含的轨迹信息具有重要的理论和应用价值.随着生活水平的日益提高,越来越多的汽车都配备了GPS设备,海量的GPS轨迹数据随之产生.为了减少车辆轨迹数据的存储空间,提高数据传输和数据分析速度,提出一种MapReduce架构下的大规模轨迹数据压缩策略.该策略首先提出一种基于综合时空特征的开放窗口轨迹数据压缩方法,再结合MapReduce并行计算模型,在各节点上并行压缩大规模轨迹数据.实验结果表明,本文提出的轨迹数据压缩策略虽然在压缩率上略有下降,但是保留了轨迹特征,减少了压缩误差,提高了压缩速度.

The vehicle trajectory information of vehicle GPS data has important theoretical and practical value. With the improvement of living standard, more and more cars are equipped with GPS devices and a large amount of GPS trajectory data are produced. In order to reduce the storage space of the vehicle trajectory data and improve the speed of data transmission and data analysis, we put forward a compression strategy of large scale trajectory data based on MapReduce architecture. Firstly, this method proposed an opening window trajectory data compression method based on the comprehensive spatial and temporal characteristics, and then using the MapReduce parallel computing model to compress trajectory data at each node. The experimental results show that the compression strategy proposed in this paper decreases slightly in compression ratio, but significantly retain the features and reduce the compression error and improve the compression speed.