基于数据流聚类挖掘算法的入侵检测系统研究  

Research on Intrusion Detection System Based on Data Stream Clustering Mining Algorithm

作　　者：梅莹莹 梁月放 MEI Ying-ying;LIANG Yue-fang(School of Computer Engineering, Anhui Sanlian University, Hefei 230601, China)

机构地区：[1]安徽三联学院计算机工程学院,安徽合肥230601

出　　处：《信阳农林学院学报》2020年第3期113-118,123,共7页Journal of Xinyang Agriculture and Forestry University

基　　金：安徽三联学院科学平台重点项目(PTZD2020014);安徽省教学研究重点项目(2018jyxm0445);安徽三联学院教研重点项目(19zlgc020).

摘　　要：为解决传统入侵检测实时性不足的问题,针对当前网络安全中处理速度快、防御能力强、实时性能高等特点,研究基于数据流挖掘与入侵检测相融合的网络安全防御技术,建立新的检测模型,设计了快速剔除孤立点的算法,提出一种改进的基于衰减滑动窗口密度的数据流聚类挖掘(ASWDStream)算法,并对该算法及其在入侵检测系统中的应用效能进行了验证。仿真结果表明该算法具有较低的运行环境要求和较高的聚类准确性,入侵检测系统表现出较高的检测率和实用性。To solve the problem of real-time performance in the conventional intrusion detection system,the characteristics of current network security were analyzed,such as high processing speed,strong defense capability,real-time performance.Then this paper studied the network security defense technology based on data stream mining and intrusion detection,and a new detection model was established.Then this paper designs a way of removing isolated points,and an improved algorithm of attenuation sliding window density-based data stream clustering mining(ASWDStream)was proposed.And the application effectiveness of the algorithm in intrusion detection system was verified.The simulation results show that the algorithm has a lower requirement for operating environment and higher clustering accuracy.The intrusion detection system shows a higher detection rate and availability.

关 键 词：信息安全 入侵检测 衰减滑动窗口 数据流挖掘 聚类算法 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]