Data Quality Test And Analysis Of Streaming Media Network

With broadband and the rapid development of computer technology, Internet networks have entered every household; while Internet-based applications growing exponentially, from the initial simple text transmission, to the transmission illustrated pages; and then from the transfer web pages to audio, video data transmission, Internet information age has become an indispensable information carrier.Video and audio data transmission and application of network information has gradually become the mainstream, Internet on the emergence of a large number of video-based sites and applications products, commonly seen Google, Youtube, Sina video, CCTV's online video, PPLive of P2P TV and so on. Streaming media data transmission are informative, high-bandwidth, real-time requirements and high, meaning that the user experience is essential! How to detect a particular streaming media network, the user experience it? Gomez's Gomez Streaming product is streaming through the network data collection, computing the relevant parameters to carry out the analysis of the user experience of a streaming media network monitoring tool. It is deployed in many global network Node, collecting streaming media monitoring data, are submitted to the central database, through a central portal to provide users with the user experience for a specific network data analysis and results display. Gomez Streaming diagnostic products for the enterprise multimedia content site experience the problem; it can automatically copy users from around the world streaming media content request and on the availability, start-up time and transmission quality of alarm problems.Each new version of the introduction of Gomez Streaming products have to go through a large number of sample space testing, test data reliable, credible and sustained and stable product quality has become critical, and the product has been distributed to a large number of commercial customers, product quality related to the company's economic interests. There are two aspects of the problems should be solved:1. With the increase in customers, Gomez Streaming products will collect more more and more data, how to ensure data quality of these huge amounts of data under the pressure of so many of our customers, and the collected data becomes more reliable from user experience?2. How about the new version of the product than the old version of the product testing of the data more reliable? The key question is, in the face of massive test data does not allow one to analyze the article. Is there a practical way?The issue is to seek to resolve the above two issues, to the mass of the test data from the statistical point of view streaming data quality analysis and research. Presents a standards-based, variance, standard deviation and other statistical theory, algorithms, and streaming media data for the indicators of the actual DNS time, Connection time, FirstByte time etc, have been analyzed and realized.This paper mainly for the following aspects of work:1. In-depth study of streaming media technology concepts and principles, streaming media performance parameters of in-depth analysis and research index.2. Combined with streaming media network platform of practical application, a detailed design based on the arithmetic mean of data quality analysis algorithm formula and process.3. Implement a data quality analysis system and seamless connection to Gomez Streaming system.4. Execute a lot of testing and analysis of the results for this data quality analysis system.