Robust AI-Based Network Intrusion Detection Under Adversarial Attacks
Supervisor Name
Mohammed Naim
Supervisor Email
mk.naim@alaqsa.edu.ps
University
Al-Aqsa University
Research field
Computer Science
Bio
Dr. Mohammed Naim is an Assistant Professor and the Head of the Faculty of Intelligent Systems and Computer Engineering at Al-Aqsa University, Palestine. He holds a Ph.D. in Networks and Telecommunications, specializing in Cryptography and Data Security. His research interests include cybersecurity, cryptography, image encryption, chaos-based cryptographic algorithms, and secure communication. He has published research articles in international scientific journals and is committed to advancing scientific research, fostering interdisciplinary collaboration, and mentoring students in innovative cybersecurity solutions.
Description
This project aims to develop and evaluate a robust artificial intelligence-based network intrusion detection system capable of identifying malicious network traffic and detecting cyberattacks. The research will investigate the vulnerabilities of machine-learning-based intrusion detection models to adversarial attacks, in which carefully manipulated network traffic attempts to evade detection. Using publicly available cybersecurity datasets, the project will implement and compare selected machine learning models, evaluate their detection performance under normal and adversarial conditions, and investigate defense mechanisms to improve their robustness. Performance will be assessed using standard metrics, including accuracy, precision, recall, F1-score, false-positive rate, and computational cost. The expected outcome is an experimentally validated intrusion detection framework that improves understanding of AI vulnerabilities in cybersecurity and provides practical recommendations for developing more reliable and resilient network security systems. The project will also provide undergraduate researchers with hands-on experience in cybersecurity, machine learning, and experimental research.