IoT intrusion detection model based on gated recurrent unit and residual network

The sample data of the existing intrusion detection models of the Internet of Things has defects such as class imbalance and insufficient feature extraction, which leads to low accuracy. Therefore, an intrusion detection model based on Gated Recurrent Unit (GRU) and Residual Network (ResNet) is proposed. Firstly, the deep convolutional generative adversarial network is used to generate a few sample data in the class imbalance data to make the sample data reach balance. Then, GRU is used to learn the data features, extract time series features of the sample data, classify the sample data features with the ResNet, and finally normalize the classification results with the softmax function. The proposed model is verified on NSL-KDD dataset, simulation experiments on NSL-KDD dataset show that the accuracy and detection rate of the proposed intrusion detection model reach 96.12% and 97.85% respectively, which are 1.86% and 2.59% higher than those of LSTM-ResNet method. And compared with GRU, LSTM and SVM, the validity of the model was verified.