Research On Cloud Storage Of Spatial Data In HBase

With the rapid development of GPS and Bei Dou Positioning Technology, the internet of things, mobile internet, vehicle networking and other things based on the application of real time positioning technology continues to mature, spatial data showing growth spurt. According to the characteristics of Internet environment, such as multi data type, fast update, high concurrent access, the traditional data management method is increasingly difficult to adapt to the demand. Therefore, how to effectively store and manage large spatial data, become the urgent problem to be solved. Cloud storage and distributed computing provides a new way to solve the above problems.Aiming at the deficiency of the current spatial data storage management method,the Hadoop open source platform and the HBase storage management are used to manage the massive spatial data.In this paper,I studied the key technologies of spatial data storage model in HBase and parallel processing, the main research results are as follows:(1) The architecture and technical principle of HDFS and HBase are analyzed, as well as the MapReduce programming model, which lays the foundation for the management of spatial data in the cloud computing environment.(2) HBase uses one-dimensional row keys to query, and spatial data by a number of two-dimensional coordinates. How to transform the two-dimensional coordinates into a one-dimensional string, has become the main issue in HBase spatial data management. This paper is based on GeoHash dimension reduction algorithm,exploring spatial data indexing method under the cloud platform.(3) The storage model of spatial data is designed based on the NoSQL characteristics of HBase. By making full use of the advantages of distributed cluster computing, a parallel spatial data query process based on MapReduce is proposed.The experimental results show that the design of the storage model, index designand parallel query process has good performance.