Network Intrusion Detection using Deep Convolution Neural Network
Vanlalruata Hnamte, Jamal Hussain · 2023
In recent years, with the rise of cyber attacks, intrusion detection systems (IDS) have become an essential component of network security. Deep learning-based approaches have shown promising results in detecting network intrusions. This study proposes a deep convolutional neural network (DCNN) approach for detecting and classifying network intrusions, and validates its performance using the InSDN dataset. The proposed DCNN model consists of two 1D convolutional neural networks connected to four fully connected layers for classifying various types of network abnormalities. Experimental results demonstrate that the proposed technique provides high precision for classification and achieved high accuracy for all DCNN models used. The proposed model was also compared to state-of-the-art intrusion detection techniques using several performance metrics, including accuracy, and loss rate. The proposed model outperformed existing methods for all parameters. The study concludes that the proposed model has the potential to enhance the security of networks and mitigate security concerns in modern life.