Research on Intrusion Detection Model Based on SSAE-ResNet
Aiming at the problems of multiple redundant features in wireless LAN traffic,increasingly prominent security,false positives and false negatives in network intrusion detection,this paper proposed an intrusion detection model combined stacked sparse encoder(SSAE)and one-dimensional residual network(ResNet).Using the wireless local area network dataset AWID as the data sample,we first preprocessed the massive high-dimensional data into data types that the model can adapt to.Then,we designed a combination model of autoencoder and residual net-work.In order to avoid overfitting of the model and incomplete feature extraction,regularization terms were added to the autoencoder,and a stacked autoencoder was designed for feature extraction.The extracted features were used as inputs to the classifier,which adopted an improved one-dimensional ResNet design to save time in converting traffic data into images.The experimental results show that this model has good results,stable running environments,which explains the effectiveness of the model.
NETL
NSTL
万方数据
