摘要
P2P及fast-flux等技术的出现使僵尸网络隐蔽性大大增强。传统人工提取特征的识别方法愈发困难并且识别精度低。该文设计一种新的基于CNN及LSTM融合网络结构,使用改进激活函数和网络结构的卷积神经网络检测空间特征,并使用长短时记忆网络检测时序特征,将两种特征并联融合用于识别僵尸网络。实验表明,该方法在精度和召回率等方面可满足僵尸网络识别需求。
The emergence of P2P and fast-flux technology makes botnet more covert.The traditional recognition method of feature extraction is more and more difficult,and the recognition accuracy is low.In order to solve the above problems,this paper designs a new fusion network structure based on CNN and LSTM.In this method,convolutional neural network with improved activation function and network structure was used to detect spatial features,and LSTM network was used to detect temporal features.Experimental results show that the method can meet the needs of Botnet identification in terms of accuracy and recall.
作者
卢法权
陈丹伟
Lu Faquan;Chen Danwei(College of Computer,College of Software,College of Network Space Security,Nanjing University of Posts and Telecommunications,Nanjing 210023,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2024年第3期328-335,共8页
Computer Applications and Software
关键词
僵尸网络
卷积神经网络
长短时记忆网络
特征并联融合
激活函数
Botnet Convolutional Neural Network(CNN)
Long and short-term memory(LSTM)
Feature parallel fusion
Activation function
