Abstract

A modern Intrusion Detection System (IDS) is an essential component of any cutting-edge information and communication technology (ICT) architecture because of the growing concern for online security and the complexity and unpredictability of cyberattacks. Integration of Deep Neural Networks (DNNs) into Intrusion Detection Systems (IDS) for increased security measures is driven by the requirement to understand the nature of these assaults, which is why their relevance grows. With a learning rate of 0.1 and 1000 epochs of execution on the 'KDD Cup-99' dataset for training and benchmarking, this article uses DNNs to anticipate assaults on Network Intrusion Detection Systems (N-IDS). Additionally, the dataset was trained using a variety of DNNs with layers ranging from 1 to 5, in order to conduct comparative analysis and determine the effectiveness. Based on the results of this study, a deep neural network (DNN) with three layers outperforms existing deep learning models and standard machine learning techniques.

Description