基于优化CS-SVM算法的DGA域名检测研究  

Research on DGA Domain Name Detection Based on Optimized CS-SVM Algorithm

作　　者：卢加奇 吕广旭 魏先燕 冯燕茹[1] 王小英[1] LU Jiaqi;LYU Guangxu;WEI Xianyan;FENG Yanru;WANG Xiaoying(Institute of Disaster Prevention,Langfang 065201,China)

机构地区：[1]防灾科技学院,河北廊坊065201

出　　处：《现代信息科技》2023年第11期77-79,共3页Modern Information Technology

基　　金：防灾科技学院教育研究与教学改革项目(JY2022B31)。

摘　　要：近年来恶意软件融合域名生成算法,生成大量的恶意域名严重威胁网络安全。目前现有的恶意域名检测方法大多都存在检测效率低等问题。提出一种通过采用优化后的布谷鸟搜索算法(CS)对支持向量机(SVM)进行优化,即BCS-SVM方法,该方法能够适应DGA域名检测场景。实验采用开放域名数据作为样本集,对文章提出的DGA域名检测方法进行训练,并通过域名向量转换、检测模型训练、参数调优,最终完成了一种较为高效的DGA域名检测模型。In recent years,malware integrates domain name to generate algorithms,and generates a large number of malicious domain names which is a serious threat to network security.Most of the existing malicious domain name detection methods have problems of low detection efficiency.This paper proposes a Support Vector Machine(SVM)optimization method by using an optimized Cuckoo Search algorithm(CS),namely the BCS-SVM method,which can be adapted to the DGA domain name detection scenario.The experiment uses open domain name data as the sample set,trains the DGA domain name detection method proposed in this paper,and an efficient DGA domain name detection model is realized through domain name vector conversion,detection model training and parameter tuning finally.

关 键 词：特征选择 DGA域名 布谷鸟搜索算法 支持向量机 

分 类 号：TP393.0[自动化与计算机技术—计算机应用技术]