Research on Intrusion Detection Technology Based on CNN-SaLSTM

Abstract As Internet-connected application devices become more and popular, services need to be done through the network, which also leads users paying attention network security performance. Due continuous iterative development of cyber attack means scale, it is difficult conduct passive detection systems such as traditional intrusion mechanisms endless attacks. Later, was studied an active defense technique compensate for shortcomings safety techniques. Active response technology has attracted researchers at home abroad. The complex, engineering large-scale scenarios presented by attacks prevent original system meet users’ needs With expansion increase traffic rapid iteration means, performance put higher requirements. Therefore, we introduced CNN, LSTM self in deep learning into invasion performed experiments tensorflow framework, increasing accuracy 97.4%.