An Enhanced Intrusion Detection and Prevention using Artificial Neural Network Technique
Kathirvel P, B. Mary Reni · 2025
Intrusion Detection Systems (IDS) play a crucial role in cyber security by identifying unauthorized access and malicious activities. However, traditional detection methods struggle to combat increasingly sophisticated cyber threats. This study explores the application of Artificial Neural Networks (ANNs) for IDS, leveraging their ability to learn from data and recognize complex network traffic patterns. ANNs are trained on labeled datasets containing common intrusion behaviors, enabling them to support both signature-based and anomaly-based detection. In particular, the system successfully identified a previously unknown attack pattern, highlighting its effectiveness. Using the ANN backpropagation algorithm, the model iteratively improves accuracy by adjusting weights based on error rates. Despite challenges such as high computational requirements, potential system overload, and the need for well-labeled training data, ANNs show significant promise in enhancing IDS capabilities against evolving cyber threats.