Research on Lightweight Intrusion Detection Model Based on Convolutional Neural Network

Network security is facing increasingly severe challenges in today's digital era. As a key technology to ensure network security, the improvement of intrusion detection system (IDS) performance is crucial. This paper proposes a lightweight network intrusion detection model based on convolutional neural network (CNN), which aims to efficiently identify malicious traffic in the network. Through data enhancement and optimized network structure, and using real-world network traffic data set CIC-IDS, the model can improve the detection accuracy while maintaining low computational resource consumption. This paper describes in detail the complete process from data collection, preprocessing, to model construction, training, evaluation and optimization, and uses a variety of evaluation indicators and visualization tools to analyze the results. The research results show the performance and advantages of the model in intrusion detection tasks, and provide a more reliable and effective solution for network security protection.