The Storage And Query Processing Of Multi-attribute Moving Objects

In recent years,with the rapid development of location service technology and mobile communication technology,location based service are getting more and more popular.The moving objects databases have attracted extensive attention,and its main responsibility is to manage spatio-temporal objects whose locations change over time in the database.Moving objects databases play an important role in traffic management system,urban planning,route recommendation and other fields.However,the existing researches on moving objects only focus on the temporal and spatial attributes,ignoring the other properties of moving objects which could be useful in practical applications,such as color and vehicle type,the moving objects' query with attributes can be widely applied in different application scenarios.In this paper,we proposed a new storage model,a new index structure and corresponding continuous distance query algorithms for multi-attribute moving objects.The main research works are as follows:(1)A nested storage model for multi-attribute moving objects is designed,and the static attributes of moving objects are added in the form of nested tables to realize the storage of multi-attribute moving objects.In order to solve the problems of data format and noise of original data,a data preprocessing algorithm is proposed,and multi-attribute moving objects are stored in the open source extensible database SECONDO.We also compared our model with the non-nested storage model using experiment and verified the effectiveness of the nested multi-attribute moving objects storage model.(2)In order to implement continuous distance query of multi-attribute moving objects,this paper introduces the 3D R-tree index with attributes and octree index with attributes,which are used to support the continuous distance query algorithm for multi-attribute moving objects.Furthermore,we optimized the filtering method to improve query efficiency.By using different parameters in the query algorithm,we compared our index with the existing moving objects indexes in designed experiments.The experiment results verified the validity of the algorithm,and the performance is improved by30%-40%.(3)In order to analyze the storage and query process of multi-attribute moving objects more intuitively,we designed a read-write monitoring system for multi-attribute moving objects database files(such as relational tables,indexes and logs).We also implemented a tool to monitor the reading and writing states of the open source extensible database SECONDO files using Linux kernel technology and Java technology.The tool can display reading and writing states of database files in the database query in the real time.It can also show more comprehensive results by analyzing historical monitoring data.