| In recent years, OTT video into millions of households, video traffic accounted for more than 80% of the overall network traffic, and is growing rapidly. But at the same time video users' complaints are increasing year by year. Users have limited network bandwidth. Due to the congestion of the network and lower transmission rate makes a lot of video users do not get a higher video viewing experience. How to analyze and locate video poor video business reasons, in order to improve the user's video viewing experience, is placed in all video service providers or video content providers in front of a major issue. In order to improve the user experience to retain more users, video service providers or video content providers urgently need to establish a comprehensive OTT video service quality assurance system.At present, there is no uniform standard for how to evaluate the performance and quality of video services. Traditional video measurement technology is mainly based on UDP RTSP / RTP real-time streaming media research, measurement indicators such as network jitter,delay factor, network packet loss and so on, However, these techniques do not apply to measurements of OTT video service performance based on HTTP / TCP transport. From another point of view, the traditional research focused on the subjective feelings of users to establish evaluation model, for this OTT video objective indicators of monitoring and indicators of access to research is almost blank, so this area has a larger research space.This paper mainly focuses on the video service quality evaluation based on OTT video service. In the client build business monitoring and analysis system, real-time monitoring of the current mainstream video sites. From the video transmission status in the network, the video itself attributes, the user broadcast behavior, screening out the more commonly used video business objective indicators. Through the signaling acquisition analysis, video decoding analysis, user behavior monitoring,measurement and estimation methods to obtain the above indicators. We design and implement a monitoring and analysis system for passive acquisition of video services. The system consists of network data capture,video decoding analysis, link analysis, information display and storage of four modules.This paper first introduces the research background and significance of the subject. The research status and existing products of OTT video service monitoring and analysis at home and abroad are described, and the content and research objectives are pointed out. Then it introduces the related technology used in the design and development process, including OTT video related technology, network data fetching technology, video codec technology and database technology, and makes the technical preparation for the development of the system. Then introduced the monitoring system design and development work, from demand analysis to summary design and then to the detailed design and code implementation, in strict accordance with the software design and development process started. In the process of system development, we have carried on the detailed research to the current mainstream video websites Youku, Letv, Fantastic Art, Tencent and Mango TV, and clarified the specific meaning of various signaling in the video business process.The method of obtaining the objective parameters of each video is pointed out, and rigorous test is carried out for each parameter in the later stage of system development. We also designed and implemented the OTT video system in a laboratory environment. First, in order to verify the feasibility of some parameters to obtain the method, the second is to be controlled through the video player to get more video-related business parameters. Experimental platform is mainly composed of video server,audio and video players in two parts, mainly through the internal monitoring interface in the player to get the target parameters. |
PDF Full Text Request