基于Hadoop的视频监控数据中心关键支撑技术研究与应用

发布时间：2018-02-01 13:08

本文关键词： Hadoop 数据中心分布式存储分布式计算视频转码　出处：《北京邮电大学》2013年硕士论文　论文类型：学位论文

【摘要】：近年来,物联网和云计算技术都在飞速发展,物联网技术的广泛应用给人们带来了巨大的便利,云计算凭借其强大的存储和计算能力,在海量数据存储和分布式计算领域的应用也越来越多。当前物联网与云计算的联系越发紧密,两者相结合也是未来的发展趋势,因此如何将云计算技术与物联网技术有效的结合,如何构建面向物联网的云数据中心以及数据中心中的计算、存储、管理等关键技术的研究和应用便成为了最近几年研究热点。本文深入分析了Hadoop的分布式存储和MapReduce计算引擎以及分布式系统管理等技术,对比了常见云数据中心的架构,包括存储和计算模型等,从视频监控系统这一物联网具体应用场景出发,设计并实现了一种以Hadoop作为核心技术的面向视频监控领域的云数据中心原型系统。论文从数据中心硬件基础平台的设计和资源整合虚拟化、分布式存储系统和计算引擎、视频监控数据中心核心服务等几个层面描述了数据中心的整体架构,在对各层中关键技术进行研究的基础上,实现了监控视频分布式采集和管理、海量监控视频文件存储、分布式视频文件处理等服务。相比传统的视频监控系统,基于云计算的视频监控系统从计算性能、存储性能等方面都将有极大的提升。本文对数据中心的核心服务进行测试和比对,实验结果表明本文提出的视频监控数据中心系统提高了多路流式视频数据收集的可靠性,显著降低了离线视频处理的时间。
[Abstract]:In recent years, the Internet of things and cloud computing technology are both rapid development, the wide application of Internet of things technology has brought great convenience to people, cloud computing with its powerful storage and computing capabilities. There are more and more applications in the field of mass data storage and distributed computing. At present, the connection between Internet of things and cloud computing is more and more close, and the combination of the two is also the future development trend. Therefore, how to effectively combine cloud computing technology with Internet of things technology, how to build cloud data center for the Internet of things and the computing and storage in the data center. The research and application of key technologies such as management has become a hot topic in recent years. This paper analyzes the distributed storage, MapReduce computing engine and distributed system management technology of Hadoop, and compares the structure of common cloud data center. Including storage and computing models, from the video surveillance system, a specific application of the Internet of things scene. This paper designs and implements a prototype system of cloud data center based on Hadoop, which is oriented to video surveillance field, and designs the hardware platform of data center and the virtualization of resource integration. Distributed storage system, computing engine, core services of video surveillance data center and so on, describe the overall structure of the data center, on the basis of the research on the key technologies in each layer. Compared with the traditional video surveillance system, the video monitoring system based on cloud computing has achieved the performance of video monitoring system based on cloud computing, which realizes distributed video collection and management, mass video file storage, distributed video file processing and other services. Storage performance will be greatly improved. This paper tests and compares the core services of the data center. The experimental results show that the proposed video surveillance data center system improves the reliability of multi-channel video data collection and significantly reduces the time of offline video processing.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP308;TN948.6

【参考文献】