Point To Point Internet Quality Evaluation And Anomaly Analysis

With the rapid development of computer technology and the high requirements of informatization, especially the social class applications, such as micro blog, becoming more and more popular, makes the Internet popularizing rate improves constantly. At the same time, people require better qualities of all kinds of service on the Internet. In order to improve the service quality and satisfy people’s use demand, monitoring the network service quality is very necessary. Through analysis historical data, it can be timely to find the existing anomaly information in the current network, and then to solve the problems, and also to improve the performance of network. Not only for ordinary Internet users, but also for service provider, It contains very important meaning.This thesis introduces the current research on Internet service quality measurement. Based on the way of measurement, it can be mainly divided into active measurement and passive measurement. And then through the analysis of the Internet service status, this thesis selects the Web (WWW) service as the main service to carry on the analysis, and introduces the definition of measurement parameters which are gained based on Web service process. So far, there is no formal standards for Web service quality evaluation, this thesis, through developing a point-to-point Internet quality monitor system in the laboratory using the active measurement way to simulation users’ process of surfing Internet, describe a way to measure target site’s service quality in real-time, so as to react the real user’s experience. Then this thesis describes the monitor system’s structure, mainly describes the center server module design. At the same time in order to analysis the anomaly data, this thesis designed the anomaly analysis module, using the clustering algorithm to analysis abnormal data.