ResInceptNet-SA: A Network Traffic Intrusion Detection Model Fusing Feature Selection and Balanced Datasets

Network intrusion detection models are vital techniques for ensuring cybersecurity. However, existing models face several challenges, such as insufficient feature extraction capabilities, dataset imbalance, and suboptimal detection accuracy. In this paper, a new type of model (ResIncepNet-SA) based on InceptionNet, Resnet, and convolutional neural networks with a self-attention mechanism was proposed to detect network intrusions. The model used the PCA-ADASYN algorithm to compress network traffic features, extract high-correlation feature datasets, and oversample and balance the feature datasets to classify abnormal network traffic. The experimental results show that the accuracy, precision, recall, and F1-score of the proposed ResIncepNet-SA model using the NSL-KDD dataset reach 0.99366, 0.99343, 0.99339, and 0.99338, respectively. This model enhances the accuracy of abnormal network traffic detection and outperforms existing models when applied to imbalanced datasets, offering a new solution for network traffic intrusion detection.