Research And Implementation Of Internet Video Streaming Identification Technology

With the rapid development of the Internet and multimedia technologies, theInternet multimedia services become a common Internet applications, especially themultimedia became a popular Internet application service making it vivid. The widerange of media services, gives us more visual enjoyment and gradually into people’slives. Video’s excellent quality of service, occupied a place in Internet services. How toprovide faster and better video service has become a hot research topic in recent years.Streaming media data has characteristics of the large amount and continuous, it oftengives network heavy load. In order to provide users better video services, this paperthrough the analysis of the detailed investigation of streaming media data transmissionmechanism, summed up the streaming media information mining method toidentification and get the information of network video stream.Through the analysis of internet transmission of video data packet, today’s Internetvideo can be classified in two different forms: P2P video streaming and Web videostreaming.P2P video streaming use P2P architecture and Web video streaming usetraditional C/S architecture, two streaming media are slightly different in transmissionprotocol. These two types of video is very popular in today’s Internet applications, P2Pvideo such as PPStream, QQLive fengxing etc, Web video such as: YouTube, Tudou,Youku etc. Through the analysis of these well-known sites, we know the transportprotocols of the two video streams. Furthermore, through the obtained data packets atrouter side to get video information such as: name, type, parameters.According to the research results, we summarized a series of video stream miningmethod. Based on the method, we develop a online video user information miningprototype system. Video stream to identify the network operator can recognize the videodata in the network and adding video service chain. Not only can improve the quality ofvideo services, but also to watch the differences in quality services and charges itpossible. It can provides the basis for personalized service, at the same time providesdifferentiated services and Fee-based services, also can statistics the online video usersinformation. This system architecture using the OpenFlow network architecture technology which has a large advantage on continuous streaming data transmission,Simulation development tools selected the OPNET. System simulates the transmissionof video streaming on the Internet, identify and storage the video stream data, analysisthe user viewing habits. According to a large number of experimental tests, the proposedmethod is true and the system of interest mining has a high level of performance andcorrectness. The result of this study provides a theoretical basis, to improve the qualityof service and network of streaming video packet scheduling improvements.