Research On Storing And Indexing AIS Data In Cloud Environment

AIS(Automatic Identification System)data plays an important role in ship monitoring,navigational aid and collision avoidance.Efficiently storing and managing AIS data is meaningful for the development of water transport and the related researches.But AIS data is sent frequently,in the case of a large number of ships,AIS data storage and management system is facing the challenges of high-frequency write,massive data storage and multi-dimensional query.Traditional relational database management systems have low performance when processing massive data.And they work ineffectively to manage the growing spatio-temporal data.Therefore,an AIS data storage model based on HBase in cloud environment is developed.In order to support efficiently multi-dimensional query,a two-tier index structure of AIS data in cloud environment is explored.At the same time,the query algorithms based on the storage model and index structure for AIS data are designed.The main contents of this thesis are as follows:(1)Based on the characteristics of AIS data,a storage model of AIS data in cloud environment is designed.In order to manage massive AIS data efficiently,the storage model of AIS data in cloud environment is designed.AIS message is decoded in parallel with Spark in the model,and AIS data is stored in HBase.Considering the characteristics of AIS data,the storage schemas of the main table and the side table are designed from the aspects of rowkey and column family.The mian table is for the complete AIS data,while the side table is for the linearized spatio-temporal data.(2)An improved algorithm of R tree which names GeoR tree is developed,and a two-tier spatio-temporal index for AIS data in cloud environment is designed.HBase is inefficient to support the multi-dimensional query on AIS data.Therefore,GeoR tree which is based on the Geohash algorithm and R tree is proposed.With the GeoR tree and the B+ tree,a two-tier spatio-temporal index,GeoRB,is designed.The upper layer uses GeoR tree to index the spatial information,and the lower layer uses the B+ tree to index the linearized spatio-temporal data.This index structure lays the foundation for the efficient query on AIS data.(3)Based on the designed storage schema and the index structure,three query algorithms for AIS data are designed.Based on the storage model,a trajectory data query algorithm for AIS data is designed.Based on GeoRB index structure,a range query algorithm for AIS data is developed.And based on the range query algorithm,a ship traffic volume query algorithm is proposed.(4)Prototype system in cloud environment is designed and experiments with the AIS data are performed.Based on the above research results,an AIS data storage and management prototype system is developed in cloud environment.Experiments from different aspects are performed on the large AIS data set collected from the Yangtze River Channel.The experimental results show that the designed storage model and index structure is suitable for storing and managing AIS data.