三网融合下广电视频内容的存储技术研究

发布时间：2018-07-23 10:24

【摘要】：随着信息技术的迅猛发展和三网融合的不断推进，相对封闭的电信网和广电网被接入互联网，，这直接带来了用户规模的爆发式增长和数据量TB级的速度增长。三网融合后网络的开放性、交互性不断提高，网络视频、视听等融合业务将获得空前的发展，在极大地丰富广大消费者文化生活的同时，也对三网融合下信息的存储与存取带来了严峻挑战。在自主开发的基于三网融合的内容监管系统HUSTRIM中，针对三网融合下的广电视频内容具有数据处理量大、宽带要求高、存储容量多、实时性强等一系列特点。进行内容的海量数据存储架构分析，探索了一种能够满足对视频监管和存储功能及性能需求的存储结构。 HUSTRIM存储子系统实现了三网融合下海量非结构化数据的存储，为其他子系统提供了高效的存储和检索支持。首先分析了传统数据存储结构的可靠性、扩展性、容错性以及构建成本；再讨论了海量数据存储技术及原理，分析海量数据对存储系统的特定功能和性能要求，研究基于Google云存储的HDFS存储结构；最后开发了一种基于Hadoop的具有高可用，高并发和高扩展性的存储结构，来满足HUSTRIM系统对存储子系统的功能和效率要求。
[Abstract]:With the rapid development of information technology and the continuous advancement of the integration of three networks, the relatively closed telecommunication network and radio and television network are connected to the Internet, which directly brings the explosive growth of the user scale and the growth of the speed of the data TB level. After the integration of the three networks, the openness and interactivity of the network will continue to improve. The network video, audio-visual and other fusion services will be developed unprecedentedly, which will greatly enrich the cultural life of the broad masses of consumers at the same time. It also brings a severe challenge to the storage and access of information under the three-network convergence. In HUSTRIM, a content supervision system based on triple network fusion, radio and television video content has a series of characteristics such as large amount of data processing, high requirement of broadband, high storage capacity, strong real-time and so on. Based on the analysis of mass data storage architecture, a storage structure that can meet the requirements of video supervision and storage function and performance is explored. The HUSTRIM storage subsystem realizes the storage of massive unstructured data under the integration of three networks. It provides efficient storage and retrieval support for other subsystems. Firstly, the reliability, extensibility, fault tolerance and construction cost of the traditional data storage structure are analyzed, and then the technology and principle of mass data storage are discussed, and the specific functions and performance requirements of the mass data storage system are analyzed. The HDFS storage structure based on Google cloud storage is studied. Finally, a high availability, high concurrency and high scalability storage structure based on Hadoop is developed to meet the functional and efficiency requirements of the storage subsystem in HUSTRIM system.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP333

【相似文献】