Doctorate thesis defense on April 05th 2025 at 09H00 AM ,in Amphitheater Ibn Khaldoun, SUP'COM 2.
Entitled : ARTIFICIAL INTELLIGENCE TECHNIQUES FOR THE DETECTION AND RESOLUTION OF CYBER-ATTACKS IN SMART CITIES
Presented by : MEHDI HOUICHI
President |
Pr. Sadok El ASMI |
SUP'COM, Tunisia |
|
Reviewers |
Pr. Mohamed MOSBAH |
University of Brdeaux, France |
|
|
Dr. Hanene BOUSSI |
University of Tunis el Manar, Tunisia |
|
Examiner |
Mme Ryma ABASSI |
ISET'COM, Tunisia |
|
Supervisor |
Pr. Adel BOUHOULA |
Arabian Gulf University, Bahrain |
|
Co-supervisor |
Pr. Mourad MNIF |
SUP'COM, Tunisia |
Smart cities represent the future of urban development, leveraging Information and Communication Technologies (ICT) and the Internet of Things (IoT) to enhance public safety, resource management, and service efficiency. Despite these benefits, smart cities face significant cybersecurity challenges. Their interconnected systems increase vulnerability, requiring robust measures to protect critical infrastructure. Smart city networks collect vast amounts of data, raising concerns about privacy and the potential for cyber-attacks. Key challenges include the complexity of securing interconnected systems, privacy concerns from data collection, and vulnerabilities like Distributed Denial of Service (DDoS) and ransomware attacks. These threats exploit weaknesses in IoT devices, potentially disrupting services like energy, transportation, and emergency response.
In this thesis, we propose a three-phase AI-based methodology to secure smart cities: (i) cyber-attack detection, (ii) tracking and localizing the attack source, and (iii) mitigating threats. This solution uses machine learning techniques such as Convolutional Neural Networks (CNNs) and Random Forest classifiers to enhance detection accuracy and minimize false positives. Extensive testing on datasets like NSL-KDD and N-BaIoT confirmed its superior performance compared to traditional Intrusion Detection Systems (IDS). The system also localizes attack sources using Deep Packet Inspection (DPI), which enables detailed traffic analysis and timely threat containment.
Finally, the system applies response strategies, such as isolating compromised devices or blocking malicious traffic, to minimize the impact on critical services. The proposed approach offers adaptive security, real-time threat detection, and a significant reduction in false positives, making it a promising solution for the evolving cybersecurity needs of smart cities.
Smart Cities, Cybersecurity, Intrusion Detection, Artificial Intelligence, IoT Security, Real-Time, Monitoring
Maintenant, allez pousser vos propres limites et réussir!