流媒体数据分发技术研究与实现
发布时间：2018-09-05 12:16
【摘要】： 随着网络技术、视音频压缩技术和计算机存储技术的发展,以及在各种网络中对流媒体服务需求的增加,流媒体技术也得到快速的发展,它在视频点播、流媒体新闻发布、网络广告、在线直播、视频会议、远程教育等流媒体信息服务领域已展现出广阔的应用前景。而多媒体内容的传输占用了当前Internet上大部分流量,导致骨干带宽紧张、用户接入时延增加和用户管理混乱。为缓解网络压力过大和满足用户需求,许多新技术被提出和应用,其中包括代理缓存技术和P2P网络模型架构。 本文深入研究现有的流媒体系统架构和代理缓存技术,研究和设计出适合于同构和异构网络下的数据分发策略,通过仿真实验验证了数据分发策略设计的合理性和有效性,最后,结合数据分发策略实现带代理缓存服务器的流媒体系统。策略包括系统拓扑架构、缓存方式、缓存算法、数据分布规划和数据替换算法。系统拓扑架构采用分层式拓扑架构,分为代理层和用户层,根据同构网络和异构网络的相异性而对系统用户层进行不同设计;缓存算法是基于等长分段缓存方式,结合媒体流行度、用户特性和网络带宽进行设计;数据替换算法是基于LFU,结合规划数据分布进行设计。
[Abstract]:With the development of network technology, video and audio compression technology and computer storage technology, as well as the increasing demand for streaming media services in various networks, streaming media technology has also been rapidly developed. Online advertising, online live broadcast, video conferencing, distance education and other streaming media information services have shown a broad application prospects. However, the transmission of multimedia content takes up most of the current Internet traffic, which leads to the shortage of backbone bandwidth, the increase of user access delay and the confusion of user management. In order to relieve the pressure on the network and meet the needs of users, many new technologies have been proposed and applied, including proxy caching technology and P2P network model architecture. In this paper, the current architecture of streaming media system and proxy cache technology are deeply studied, and the data distribution strategy suitable for isomorphic and heterogeneous networks is studied and designed. The rationality and effectiveness of the design of data distribution strategy are verified by simulation experiments. Finally, the streaming media system with proxy cache server is implemented with data distribution strategy. Strategies include system topology, caching, caching algorithms, data distribution planning and data replacement algorithms. The system topology adopts hierarchical topology architecture, which is divided into proxy layer and user layer. According to the difference between isomorphic network and heterogeneous network, the system user layer is designed differently. The data replacement algorithm is designed based on LFU, and planning data distribution, which is based on media popularity, user characteristics and network bandwidth.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：TN919.8
本文编号：2224241
[Abstract]:With the development of network technology, video and audio compression technology and computer storage technology, as well as the increasing demand for streaming media services in various networks, streaming media technology has also been rapidly developed. Online advertising, online live broadcast, video conferencing, distance education and other streaming media information services have shown a broad application prospects. However, the transmission of multimedia content takes up most of the current Internet traffic, which leads to the shortage of backbone bandwidth, the increase of user access delay and the confusion of user management. In order to relieve the pressure on the network and meet the needs of users, many new technologies have been proposed and applied, including proxy caching technology and P2P network model architecture. In this paper, the current architecture of streaming media system and proxy cache technology are deeply studied, and the data distribution strategy suitable for isomorphic and heterogeneous networks is studied and designed. The rationality and effectiveness of the design of data distribution strategy are verified by simulation experiments. Finally, the streaming media system with proxy cache server is implemented with data distribution strategy. Strategies include system topology, caching, caching algorithms, data distribution planning and data replacement algorithms. The system topology adopts hierarchical topology architecture, which is divided into proxy layer and user layer. According to the difference between isomorphic network and heterogeneous network, the system user layer is designed differently. The data replacement algorithm is designed based on LFU, and planning data distribution, which is based on media popularity, user characteristics and network bandwidth.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：TN919.8
【引证文献】
相关期刊论文 前1条
1 张祖礼;何洪波;尹海龙;;基于多应用服务模式的视频转换中间件的开发和实现[J];微电子学与计算机;2013年02期
，本文编号：2224241
本文链接：https://www.wllwen.com/wenyilunwen/guanggaoshejilunwen/2224241.html
