@inbook{0b0fc89e161c44f8960ab71b26e3718b,
title = "Network Attack Detection on IoT Devices Using 2D-CNN Models",
abstract = "The rapid development of IoT networks emphasises the critical importance of robust security measures. Consequently, anomaly-based intrusion detection systems using machine learning techniques have garnered significant attention due to their ability to detect unseen attacks. This study introduces neural network approaches for network attack detection. We propose supervised learning approaches, combining Artificial Neural Networks (ANN) and 2D Convolutional Neural Networks (2D-CNN) to detect attacks on the IoT-23 dataset. We only consider packets that belong to IPv4 and one of the three protocols: TCP, UDP, or ICMP. The ANN and 2D-CNN have achieved the highest accuracy of 99.71\% and 99.34\% on the IoT-23 datasets, respectively. Furthermore, by looking at the packet level, the 2D-CNN models show an approximately 40\% improvement in feature extraction time compared to ANN models. Our approach offers innovative solutions for network attack detection systems which can be mapped on the latest computing architectures, including CNN accelerators and FPGAs.",
keywords = "2D-CNN, ANN, IoT devices, Network security",
author = "Ngo, \{Duc Minh\} and Dominic Lightbody and Andriy Temko and Cuong Pham-Quoc and Tran, \{Ngoc Thinh\} and Murphy, \{Colin C.\} and Emanuel Popovici",
note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.",
year = "2023",
doi = "10.1007/978-3-031-46749-3\_23",
language = "English",
series = "Lecture Notes on Data Engineering and Communications Technologies",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "237--247",
booktitle = "Lecture Notes on Data Engineering and Communications Technologies",
address = "Germany",
}