Degree

Master of Science in Data Science

Department

Department of Computer Science

Faculty/ School

School of Mathematics and Computer Science (SMCS)

Date of Submission

Fall 2024

Supervisor

Muhammad Zain Uddin, Lecturer Computer Science-SMCS

Keywords

Internet of Things (IoT), Machine Learning, Anomaly Detection, XGBoost Classifier, Network Security

Abstract

Internet of Things (IoT) devices exponentially grew making it possible for an unprecedented level of connectivity and automation revolutionizing the personal and industrial landscapes. But it’s a surge in IoT adoption that has also increased security vulnerabilities, making networks devices that easily become targets for cyberattacks, unauthorized access and data breaches. To address these critical challenges this project presents a dual layered Layered Machine Learning Framework for IoT Device Security, which fortifies IoT networks in a dual layered way. The first layer equips users with manual classification tools so they can classify devices as Trusted, Blocked or Unknown, thereby establishing a foundational security perimeter that allows only trusted devices to access the network. Second Layer utilizes an XGBoost Classifier to continuously perceive the behaviour of the Trusted devices and use Machine Learning techniques to detect and respond to abnormal behaviours that may be indicative of the occurrence of a breach in real time. The framework also includes an easy-to-use airplane application interface for airplanes and router APIs to help manage network traffic in a granular and dynamic site and device manner. The framework is shown, in preliminary form, as ways to make network security better and allow device connectivity. Beyond this system, future work will include deep learning-based monitoring to further enhance the system’s ability to detect anomalies. This all-in-one security solution not only addresses the current vulnerabilities of IoT ecosystems but also provides a scalable technology-oriented model based on the changing landscape of the world that adds a substantial contribution to the area of network security.

Document Type

Restricted Access

Submission Type

Research Project

Download

The full text of this document is only accessible to authorized users.

Access content accompanying this publication

Share

COinS