Enhancing network security: an intrusion detection system using residual network-based convolutional neural network

Enhancing network security: an intrusion detection system using residual network-based convolutional neural network

연구 분야: Networking

학회: Cluster Computing

With the rapid growth of cyber threats, traditional Intrusion Detection Systems (IDS) face limitations in accurately detecting evolving and diverse attack types. These systems often struggle with scalability, manual feature extraction, and generalization to new attack patterns. To address these issues, this study proposes a Residual Network-based Convolutional Neural Network (ResNet-CNN) model to enhance the detection and classification of network intrusions. The proposed model combines advanced deep learning techniques for automatic feature extraction and robust classification performance. Using the NSL-KDD dataset, five experimental setups were conducted, including binary classification, binary classification with feature division, multiclass classification, multiclass classification with feature division, and binary classification of attack types. The proposed model achieved remarkable accuracies of 98.94% and 98.92% for binary and multiclass classification tasks, respectively. Additionally, the model demonstrated exceptional accuracy in detecting individual attack types, exceeding 99%. This research highlights the significance of leveraging basic features for optimal classification and introduces a promising methodology for safeguarding against sophisticated cyber threats.