Zero-day detection on IOT networks

• Main Author: Oh, Jia Sheng
• Format: Final Year Project / Dissertation / Thesis
• Published: 2024
• Subjects: HC Economic History and Conditions; T Technology (General); TD Environmental technology. Sanitary engineering
• Online Access: http://eprints.utar.edu.my/6670/1/fyp_CS_2024_OJS.pdf; http://eprints.utar.edu.my/6670/

This project aims to develop a federated learning-based solution for detecting zero-day attacks on IoT devices. Zero-day attacks exploit vulnerabilities unknown to developers or security experts, making them difficult to detect and prevent using traditional security measures. The project's main objectives are to leverage distributed machine learning techniques to train models on data stored on different devices without transferring the data to a central server, improve early detection of zero-day attacks, and reduce network traffic. By detecting and addressing zero-day attacks faster, IoT device companies can minimize the potential impact of such attacks and prevent further vulnerability exploitation. Users with IoT devices vulnerable to such attacks risk having their personal information stolen, hijacked, or becoming victims of cyberattacks. Rapid detection and response to zero-day attacks can help minimize these risks and protect users' privacy and security. The project's impact and significance include improving user privacy, reducing network traffic, increasing the precision of zero-day attack detection models, and providing quicker responses for identifying IoT device zero-day threats.