分布式协调服务的室内定位中间件研究  

Research for Indoor Localization Middleware with Distributed Coordination Service

作　　者：尹书成[1,2] 黄刘生[1,2] 朱振宇[1,2] 徐宏力[1,2] 

机构地区：[1]中国科学技术大学计算机科学与技术学院,合肥230027 [2]中国科学技术大学苏州研究院,江苏苏州215123

出　　处：《小型微型计算机系统》2017年第4期781-785,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金联合基金项目(U1301256)资助;国家自然科学基金面上项目(61272133)资助;国家发改委物联网专项基金项目(2012-2766)资助

摘　　要：随着智能手机的普及,室内定位伴随着移动互联网浪潮开始在智能家居、智慧商城和公共安全应急响应等应用场景扮演着日益重要的角色.然而室内定位系统有着数据量大,并发性强,大数据分析等传统的单节点服务器无法支撑的需求,因此本文采用大型分布式架构的思想,提出一种基于组件的分布式室内定位中间件架构.该中间件内部耦合性低,可以根据具体的室内场景大小进行灵活的弹性伸缩,不仅能够支撑大规模、实时性强的数据吞吐,而且提供大数据处理的平台,此外内部还有分布式监控组件以保障中间件安全.实验证明相较于单节点服务器,该中间件的数据读写速率、请求并发性能和大数据任务处理效率都有显著的提升.With the popularity of smartphones, indoor localization plays a more important role in smart home, wisdom mall and public safety response. However, for indoor localization, there is a need of processing large amount of data, responding concurrency strong re- quests and big-data task, which traditional single-node server can＇t support. So this paper proposes a distributed middleware based on component according to distributed architecture, which can stretch the cluster scale by the size of indoor scenes since low coupling. The middleware can not only support I/O for a large scale of real-time date,but also provide platform for big data processing. What＇s more, it has monitor component for users watching its safety. Experiments show that compared to single-node server, the distributed middleware has significant improvement on data I/O rates ,performance for concurrency and efficiency of big-data task.

关 键 词：分布式 中间件 室内定位 HADOOP 大数据 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]