基于NoSQL数据库的空间大数据分布式存储策略

发布时间：2018-03-20 00:00

本文选题：空间数据库　切入点：大数据　出处：《武汉大学学报(信息科学版)》2017年02期 　论文类型：期刊论文

【摘要】：基于关系型数据库的空间数据存储与处理是地理信息系统(geographic information system,GIS)领域的主流模式,但伴随着物联网、移动互联网、云计算及空间数据采集技术的发展,空间数据已从海量特征转变为大数据特征,对空间数据的存储和管理在数据量和处理模式上提出了新的挑战。首先分析了基于传统的集中式存储与管理模式在处理和应用大数据方面的局限性,包括存储对象的适应性、存储能力的可扩展性及高并发处理能力要求;然后在分析当前几大主流NoSQL数据库特点的基础上,指出了空间大数据基于NoSQL数据库的单一存储模式在数据操作方式、查询方式和数据高效管理方面存在的局限性;最后结合GIS领域空间大数据存储对数据库存储能力的可扩展性及数据处理和访问的高并发要求,提出基于内存数据库和NoSQL数据库的空间大数据分布式存储与综合处理策略,并开发了原型系统对提出的存储策略进行可行性和有效性进行了验证。
[Abstract]:The spatial data storage and processing of relational database based on geographic information system (geographic information system, GIS) the mainstream mode field, but with the Internet, mobile Internet, cloud computing and data acquisition technology development space, spatial data has changed from the mass characteristics for large data characteristics, storage and management of spatial data put forward the new challenge in the amount of data and processing mode. Firstly analyzes the centralized storage and management mode based on the traditional processing and application in terms of data limitations, including storage object adaptation, storage capacity scalability and high concurrent processing capability; then based on the analysis of the current several mainstream NoSQL the characteristics of the database, it points out that the big data storage model based on single NoSQL database in the data operation, query and data management efficiency The limitations of the high concurrent with GIS; finally the field of big data storage space of database storage capacity scalability and data processing and access requirements, the space data of distributed memory database and NoSQL database and comprehensive treatment strategies based on, and develop a prototype system of the feasibility and effectiveness of the verification of the storage strategy put forward.

【作者单位】：中国科学院地理科学与资源研究所;环境保护部卫星环境应用中心;北京超图软件股份有限公司;
【基金】：国家自然科学青年基金(51309210)~~
【分类号】：P208

【相似文献】