【摘要】 入侵检测系统通过分析网络流量来学习正常和异常行为,并能够检测到未知的攻击。一个入侵检测系统的性能高度依赖于特征的设计,而针对不同入侵的特征设计则是一个很复杂的问题。因此,提出了一种基于深度学习检测僵尸网络的系统。该系统利用卷积神经网络(Convolutional Neural Network,CNN)和长短期记忆网络(Long Short-Term Memory,LSTM)分别学习网络流量的空间特征和时序特征,而特征学习的整个过程由深度神经网络自动完成,不依赖于人工设计特征。实验结果表明,该系统在僵尸网络检测方面具有良好的表现。更多还原

【Abstract】 Intrusion detection systems learn normal and abnormal behavior by analyzing network traffic and are able to detect unknown attacks. The performance of an intrusion detection system is highly dependent on the design of features, and the design of features for different intrusions is a very complex issue. Therefore, a system for detecting botnets based on deep learning is proposed. The system uses Convolutional Neural Network(CNN) and Long Short-Term Memory(LSTM) to learn the spatial and temporal characteristics of network traffic respectively. The entire process of feature learning is automatically completed by deep neural networks, Does not rely on artificial design features. Experimental results indicate that the system has a good performance in botnet detection.更多还原