文献类型：期刊文献

中文题名：基于非平稳信号时频分析的DDoS攻击检测仿真

英文题名：Research on-line Monitoring method of noise pollution in Robot Master-Slave Control system

作者：李亚利[1];刘佳[2]

第一作者：李亚利

机构：[1]北京联合大学应用科技学院,北京100101;[2]北京联合大学教务处,北京100101

第一机构：北京联合大学应用科技学院

年份：2021

卷号：38

期号：5

起止页码：353-356

中文期刊名：计算机仿真

外文期刊名：Computer Simulation

收录：CSTPCD;;北大核心:【北大核心2020】；

语种：中文

中文关键词：非平稳信号;时频分析;检测

外文关键词：Non-stationary signal;Time-frequency analysis;Detection

摘要：针对传统的DDoS网络攻击检测方法未提取信号时频特征,而存在检测速度慢,甚至不能完全检测出是否受到攻击的问题,提出一种基于非平稳信号时频分析的DDoS攻击检测方法。因为攻击特征通常表现为一组非平稳的宽带信号,所以需要先构建DDoS攻击信号模型,采用包络延扩展方法,对其进行采样完成模型构建,同时因DDoS攻击非平稳宽带信号时频特征存在较好的时频聚集性,所以需要将DDoS攻击信号转变为复信号,从而提取时频特征,最后利用攻击信号模型模型与提取的时频特征货获取训练样本,计算其分类权重,完成对DDoS攻击进行检测。仿真结果证明,方法对DDoS攻击的检测速度较快,正确率高,具有良好的检测效果。
Traditionally,the DDoS network attack detection method does not extract the time-frequency features of the signal,and the detection speed is low.Therefore,a method to detect DDoS attacks based on time-frequency analysis of non-stationary signals was presented.Generally,this attack feature was presented as a group of non-stationary broadband signals,it was necessary to construct a DDoS attack signal model.And then,the envelope extension method was adopted for sampling and thus to complete the model construction.Meanwhile,the time-frequency features of non-stationary broadband signals attacked by DDoS had better time-frequency aggregation,so it was necessary to transform DDoS attack signals into complex signals,so as to extract the time-frequency feature.Finally,the model of attack signal and the extracted time-frequency feature were used to obtain training samples and calculate classification weights.Thus,the detection of DDoS attacks was completed.Simulation results show that the proposed method can detect DDoS attacks quickly and accurately.In addition,this method has a good detection effect.