基于互联网应用数据的网络性能监测及问题定位的研究

随着互联网业务的复杂性和多样性发展，传统主流的网管设备监控和探针拨测方式已不能有效支撑业务端到端的质量分析，更无法反映客户真实的业务体验和网络问题的具体定位。为了解决传统主流监测方式存在的问题，本文提出了一种基于互联网应用数据的网络性能监测及问题定位方法，通主流的互联网应用（如139邮箱）中获取海量用户的体验感知和数据链路信息，运用皮尔森相关性算法和矩阵解方程算法，将存在性能劣化的IP段进行定位，指导运维专员进行快速、准确的处理。

The research on network perfomance moitoring and the location of problem base on the data of Internet

With the complexity and diversity of Internet business development, the network management monitoring and the probe-type monitoring can not complete the analysis of network quality. It also fail to reflect the experience of user. To solve the above problems, this paper proposes the research on network perfomance moitoring and the location of problem base on the data of Internet. This scheme use Software Development Kit (SDK) to obtain the imformation of user and the information of data link in the 139mail app. Finally, Pearson correlation algorithm and matrix solution algorithm are used to locate IP with degraded performance. The output award directs the operation and maintenance specialist for quick and accurate processing.