Research And Implementation Of Streaming Media Caching Proxy Server On Multimedia Color Ring Video System

With the rapid development and popularization of the 3G mobile communication network, more and more users begin to focus on the mobile multimedia information, they not only request more enrich content of the information, but also need to get it real-time. However, there are two problems of the traditional streaming media client/ server structure:large server-side load during visit peak hours and high latency of video play. The main method solving these two problems currently is to introduce caching proxy server.Multimedia color ring video system is a multimedia interactive system which supports video on demand and video push feature for mobile users. In this paper, based on the user access features, video popularity and the media data these three basic factors, the streaming media caching proxy server used in multimedia color ring video system is designed and implemented. The prime tasks include four aspects as follows:1. On the basis of the functional analysis of multimedia color ring video system, propose the overall design of streaming media caching proxy server. The design can be divided to three modules according to the demand:the transmission control protocol module, video on demand module and video push module.2. After in-depth study on streaming transmission control protocol, propose optimal design of the protocol based on requirements of the two video modules operated in different actual situations, and implement it. According to the cache hit or miss of the two video modules, design different reply messages to transmit media data.3. For video on demand module, based on user access features and media data characteristics, propose prefix caching strategy based on video internal popularity and the value of video-based replacement strategy to design and implement.4. Video push module is the core of the system. In order to achieve low latency requirements, design and implement video full cache strategy and replacement strategy based on the video playback statistics which is a low complexity algorithm. Propose group users'video push design ideas referring to current CRBT, and implement it.