基于大数据流的网络流量检测与分析  

Network traffic detection and analysis based on big data flow

作　　者：程伟华 赵军 吴鹏 

机构地区：[1]江苏电力信息技术有限公司,江苏南京210024 [2]国网江苏省电力公司,江苏南京210029

出　　处：《南京理工大学学报》2017年第3期294-300,共7页Journal of Nanjing University of Science and Technology

基　　金：国网江苏省电力公司科技项目(SGJSXTOOYJYJ1588925)

摘　　要：针对网络流量异常检测问题,该文提出1种新的网络流量检测和分析系统。采用分布式流式处理机制达到实时检测。利用大数据平台分布式存储、数据计算分析的能力,实现网络数据分布式存储,训练网络数据协议特征库。在江苏省电力公司的营销、运行与调度等业务场景中,该网络流量检测与分析系统取得了很好的实际效果,为各个业务场景的分析提供了业务支撑。A new network traffic detection and analysis system is proposed for network traffic anomaly detection problem. A distributed stream processing mechanism is used to achieve a real-time detection ability. Network data distributed storage is achieved and a network protocol feature library is trained by using the distributed storage and the data computational analysis ability of a big data platform. The network system of detection and analysis gains a good performance in the business of marketing, operation and dispatching in Jiangsu Electric Power Company, and provides a good support for the analysis of various business scenarios.

关 键 词：数据包分析 异常检测 大数据流 网络流量检测 分布式流式处理机制 大数据平台 分布式存储 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]