摘要
室内定位系统是一种利用WIFI网络来实现标签的定位,可以在已有的WIFI网络或无线局域网上实现定位引擎,在智能家居,人力资源管理等方面开始得到广泛的应用。由于室内定位系统的场景中原始定位数据非常庞大,采用传统的关系型数据库架构无法支撑其性能需求。因此采用分布式架构的思想,设计并实现了一个数据处理中间件,承担中间件消息调度,数据库交互处理等任务。并在数据持久层提供读写分离,单库分表等工作模式,将数据按照一定规则分散到不同的集群、分区、分表中,极大降低了数据库的访问压力,一定程度上克服传统关系型数据库在处理大规模数据时的性能瓶颈。同时,为了降低第三方应用开发的复杂度,提高系统响应速度,提供基于Map Reduce的大数据分析模块。
WIFI indoor positioning system is one that utilizes a wireless network or a wireless LAN to run a positioning engine to locate tagged objects,which is being used widely in smart home,human resource management. The massive amount of raw positioning data which traditional relational database architecture can＇t meet the requirements of performance.So this paper proposes a middleware used in data process based on distributed architecture,which provides the data service, such as engine-application message scheduling and database interaction. To better process the massive amount of data produced by the indoor positioning engine, the middleware distributes the data to different routers,partitions or tables using methods such as read-write splitting and intra-database table partitioning. This makes database access significantly easier, and to some extent overcomes the performance constraints of traditional relational databases in large-scale data processing. The middleware provides module based on Map Reduce for big data analysis,in order to make the third-pary application significantly simple and optimize system performance.
出处
《电子技术（上海）》
2017年第7期62-68,共7页
Electronic Technology
基金
国家自然科学基金联合基金项目(U1301256)资助
国家发改委物联网专项基金项目(2012-2766)资助
关键词
室内定位
中间件
读写分离
单库分表
indoor position
middleware
read-write splitting
intra-database table partitioning
