数据中心视频流传输与存储技术的研究与实现

发布时间：2018-07-06 15:37

本文选题：视频监控 + Hadoop　；参考：《北京邮电大学》2012年硕士论文

【摘要】：基于云计算平台的应用,能够处理海量数据,为用户提供更可靠更高效的服务,因而得到了快速发展。如何将视频监控系统与云计算平台结合起来,并能够可靠地保存海量视频数据,为海量用户提供视频检索和播放的服务,是本文研究的重点。本文分析了传统的分布式数据中心架构,通过对视频流传输和存储两方面的优化设计,提出了基于云计算平台的统一分布式数据中心架构。在视频流传输部分,我们研究并设计了流量平衡器,采集模块,缓冲区和线程池,这些模块提高了系统的整体服务质量；在视频流存储部分,我们研究并设计了基于负载均衡的副本调整策略和文件置换机制,这些设计使系统在有限资源的情况下,能够提供更可靠的服务。本文针对数据中心中的视频流传输与存储技术实现,在实验环境下进行了测试,测试结果表明,该方法能够可靠地保存数据,高效地为用户提供点播服务,可用于大规模视频监控的数据中心设计。
[Abstract]:Applications based on cloud computing platform can deal with massive data and provide users with more reliable and efficient services. How to combine the video surveillance system with cloud computing platform, and to store the massive video data reliably, and to provide the service of video retrieval and playback for the massive users, is the focus of this paper. This paper analyzes the traditional distributed data center architecture and proposes a unified distributed data center architecture based on cloud computing platform through the optimization design of video stream transmission and storage. In the part of video stream transmission, we study and design flow balancer, acquisition module, buffer and thread pool, these modules improve the overall quality of service of the system. We study and design replica adjustment strategies and file replacement mechanisms based on load balancing, which enable the system to provide more reliable services with limited resources. Based on the realization of video stream transmission and storage technology in data center, the test results show that the method can save data reliably and efficiently provide users with on-demand service. Data center design for large-scale video surveillance.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP308;TP333

【相似文献】