基于混沌免疫聚类的网络信息流异常检测方法  

Anomaly Detection Method of Network Information Flow Based on Chaotic Immune Clustering

作　　者：胡艳华 张春玉[2] 崔亚楠 韦灵 HU Yan-hua;ZHANG Chun-yu;CUI Ya-nan;WEI Ling(College of Information Science and Engineering,Liuzhou Institute of Technology,Liuzhou Guangxi 545616,China;College of Information Engineering,Xizang Minzu University,Xianyang Shaanxi 712082,China)

机构地区：[1]柳州工学院信息科学与工程学院,广西柳州545616 [2]西藏民族大学信息工程学院,陕西咸阳712082

出　　处：《计算机仿真》2024年第8期400-403,496,共5页Computer Simulation

基　　金：国家自然科学基金项目(62262062);2021年度广西高校中青年教师科研基础能力提升项目(2021KY1710);2021年度柳州工学院科学基金项目资助(2021KXJJ08)。

摘　　要：为了保证网络安全运行,维持高质量数据储存环境,提出了基于混沌算法的无线网络信息流异常检测方法。根据无线网络异常因素划分信息流异常类型,根据不同信息流异常类型的特点建立无线网络分布模型。基于上述模型将重叠维数与异常信息流相关维度聚类,获得无线网络维度数据。利用异常因素法对已划分的聚类进行属性判定,初始化原始数据和参变量,标注集合内聚簇类别,通过不同阈值实行混沌免疫聚类算法计算,完成网络信息流异常检测。通过实验证明所提方法能够精准检测出异常信息流,且通信开销较低,可保证网络环境安全。In order to ensure the safe operation of the network and maintain a high-quality data storage environment,this paper proposed a method of detecting anomalies of network wireless network information flow based on the chaos algorithm.Firstly,the types of information flow abnormalities were classified according to the abnormal factors in wireless networks.Then,a wireless network distribution model was built according to the characteristics of different types of abnormalities.Based on the model,the overlapping dimension and the relevant dimension of abnormal information flow were clustered to obtain the dimension data of the wireless network.Moreover,the attribute judgment of the clusters was carried out by the abnormal factor method.Meanwhile,the original data and parameter variables were initialized.The clustering category in the set was marked.Finally,the chaos immune clustering algorithm was achieved through different threshold values.Thus,we completed the abnormal detection of network information flow.Experiments show that the proposed method can accurately detect abnormal information flow with low communication overhead,thus ensuring the safety of the network environment.

关 键 词：混沌算法 无线网络 网络信息流异常 维度获取 聚类融合 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]