微型电脑应用2025，Vol.41Issue(2)：29-33,5.
基于EEMD-LSTM和知识图谱的网络攻击源检测和定位方法
A Network Attack Source Detection and Localization Method Based on EEMD-LSTM and Knowledge Map
摘要
Abstract
In order to find and deal with network attacks in time,this paper proposes a network attack source detection and lo-calization method based on ensemble empirical mode decomposition long-and short-term memory(EEMD-LSTM)and knowl-edge map.By preprocessing the network traffic data,effective features are extracted and normalized to prepare data for subse-quent modeling.The EEMD-LSTM network attack source detection model is constructed,the time series data are decomposed by EEMD technology,and the attack source is detected by LSTM neural network so as to identify the abnormal behavior of the network.The knowledge map is used to locate the network attack source.By establishing the relationship map between enti-ties,the network topology is analyzed to reveal the attacker's behavior path and attack source and improve the accuracy and ef-ficiency of attack source location.Experimental results show that the proposed method has false positive rate of 1.002,false negative rate of 1.734,coverage rate of 97.1％and robustness of 4.1.This proves that the proposed method can promote the innovation and development of network security technology and maintain the order of cyberspace.关键词
集合经验模态分解/长短期记忆神经网络/网络攻击源/知识图谱Key words
ensemble empirical mode decomposition/long-and short-term memory neural network/network attack source/knowledge map分类
信息技术与安全科学引用本文复制引用
景延嵘,李宗容,李楠芳,李香,曹海山,严丽珺..基于EEMD-LSTM和知识图谱的网络攻击源检测和定位方法[J].微型电脑应用,2025,41(2):29-33,5.基金项目
国网青海省电力公司科技项目(522807230008) （522807230008）
