A Machine Learning Approach to Detect Anomalies on Edge Devices

Shamas, Awais (2024) A Machine Learning Approach to Detect Anomalies on Edge Devices. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (623kB) | Preview

Abstract

Over the years, Internet of Things (IoT) devices are increasing rapidly and is expected to extend over 20 billion by 2030. The growth underscores the increasing demand for vigorous security solutions to secure these devices from cyber-attacks. Anomalies in IoT data can lead to failure and unexpected behavior in a system. Therefore, it is crucial to detect anomalies to achieve system performance and reliability. Detecting anomalies in resources constraint devices, such as IoT devices, cause some challenges. In this research, we proposed machine learning based anomaly detection system implemented for edge device such as Raspberry Pi. Our research uses TensorFlow Lite that helps in developing a compressed model that detects malicious activities in real-time resource usage processes without demanding high computational resources. An autoencoder-based model was implemented to detect anomalies in resource usage processes. Models were trained on high performance devices and were further deployed on resource constrained hardware such as Raspberry Pi Zero 2W. The real-time inference happens every 5 seconds indicating highly accurate and timely anomaly detection for both full and compressed models by achieving an accuracy of 97.00% showcasing that lightweight models can outperform full models on the resource overhead. The key contribution of our research is the development of lightweight, scalable model for protecting the fast-growing IoT device ecosystem, which should identify new threats efficiently while assuring effective anomaly detection in resource-constraint environments.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Tomer, Vikas 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 T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Telecommunications > Computer networks > Internet of things Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	04 Sep 2025 14:44
Last Modified:	04 Sep 2025 14:44
URI:	https://norma.ncirl.ie/id/eprint/8801

Actions (login required)

View Item