Abstract—The Internet of Things (IoT) is widely adopted across numerous industries including healthcare, industrial automation, and smart infrastructure. These deployments reveal critical security vulnerabilities in systems that rely on IoT technologies, IoT networks are heterogeneous, resource-constrained, and exposed to dynamic attack vectors making them increasingly vulnerable to threats such as Distributed Denial-of-Service (DDoS) attacks, malware propagation and data breaches. Traditional Intrusion Detection Systems (IDSs) which rely on static rule-based approaches, fail to scale effectively with the diverse attack patterns and computational constraints inherent in IoT ecosystems. This paper proposes a framework that integrates Convolutional Neural Network (CNN)-based spatial feature extraction, LSTM-based temporal modeling, and an adaptive weighted ensemble for attention-guided intrusion detection. An end–to–end architecture, specifically designed to operate within heterogeneous and resource-constrained IoT environments. for the holistic spatial-temporal dependency analysis. The framework is evaluated on benchmark datasets and achieves competitive performance, attaining 98.9% accuracy, a 98.3% F1−Score, and a 0.994 Receiver Operating Characteristic-Area Under the Curve (AUC-ROC) on the IoT-23 dataset. The framework also demonstrates strong effectiveness against stealthy attacks, achieving detection rates of 99.1% for low-and-slow DDoS attacks and 98.4% for botnet attacks while adapting effectively to heterogeneous IoT environments. Because training requires substantial computation, it is performed offline in the cloud. The edge model deployed at the devices executes inference only, so it will produce a low latency of 4.2 milliseconds per sample. As well as having a low memory footprint, this means that it supports real-time operations using IoT gateways.

Index Terms—deep learning IDS, internet of things security, intrusion detection system

Cite: Ashraf Al Sharah, Tareq A. Alawneh, Anas Quteishat, Yazeed Alsarhan, Sara A. Khalil, and Mutsam A. Jarajreh, "Enhancing Cybersecurity in Internet of Things Networks Using Advanced Deep Learning Techniques for Intrusion Detection," International Journal of Electrical and Electronic Engineering & Telecommunications, vol. 15, no. 3, pp. 159-170, 2026. doi: 10.18178/ijeetc.15.3.159-170

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).