Adaptive Network Intrusion Detection Using Deep Reinforcement Learning

Javvaji, Vishnu Kumar (2024) Adaptive Network Intrusion Detection Using Deep Reinforcement Learning. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (605kB) | Preview
Preview	PDF (Configuration Manual) Download (412kB) | Preview

Abstract

In order to ensure cybersecurity by detecting and identifying accurately the malicious activities in network traffic, there is a need for an adaptive Network Intrusion Detection System (NIDS) that utilises Deep Reinforcement Learning (DRL). To create a well based model for the training of NIDS, many datasets such as UNSW-NB15 and CICIDS 2017 are used in this project. Data Sanitization, Feature Engineering, normalisation and synthetic data generation techniques like SMOTE are some of the most important step involving here. In parallel to the development and training of a Deep Q-Network (DQN) model, experience replay and epsilon decay are being used for better performance optimisation as well as stability. To provide insight into this potential, the efficiency of DQN model is examined in comparison with a benchmark Logistic Regression model on an experimental data set. Results show that the proposed approach yields a significantly better detection accuracy and robustness. This paper shows that deep reinforcement learning can be used to design a adaptive intelligent NIDS on the network level, which goes hand in hand with an active protection against rapidly changing cyber threats. Future work will focus on fine tuning this model and exploring other reinforcement learning approaches for enhanced detection.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Haque, Rejwanul UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software > Computer Security T Technology > T Technology (General) > Information Technology > Computer software > Computer Security Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	18 Jun 2025 11:57
Last Modified:	18 Jun 2025 11:57
URI:	https://norma.ncirl.ie/id/eprint/7914

Actions (login required)

View Item