摘要
随着智能电网建设的加强,电力信息网络及其承载的业务系统得到迅猛发展,网络业务流量的检测和预警具有重要的安全意义。针对目前电力信息网络缺乏处理流量异常问题的有效技术手段,提出了一种基于大数据的电力信息网络流量异常检测机制,并通过对改进的局部异常因子(M-LOF)和支持向量域数据描述(SVDD)两种常用异常检测算法的对比分析,总结出适合电力信息网络的流量异常检测方法 。
With the construction of smart grid, the electric power information network and its business system get rapid development. The early flow anomaly detection and warning are significant to the safety of network. Due to the lack of efficient measuring means to handle the flow abnormal problems, a flow anomaly detection mechanism based on big data for the electric power information network was proposed. Through the comparative analysis of two common anomaly detection algorithms, the improved local outlier factor algorithm（M-LOF） and the support vector data description（SVDD） algorithm, the suitable flow anomaly detection method for electric power information network was summarized.
出处
《电信科学》
北大核心
2017年第3期134-141,共8页
Telecommunications Science
关键词
电力信息网络
流量异常检测
局部异常因子
支持向量域数据描述
electric power information network
flow anomaly detection
local outlier factor
support vector data description
