Authors:
Deepa Parasar, R. Steffi, R. Regin, K. Daniel Jasper
Addresses:
Department of Computer Science Engineering, Amity School of Engineering and Technology, Amity University, Mumbai, Maharashtra, India. Department of Electronics and Communication, Vins Christian College of Engineering, Nagercoil, Tamil Nadu, India. School of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. School of Electronics, Electrical Engineering and Computer Science, Queens University Belfast, Belfast, Northern Ireland, United Kingdom.
Abstract:
Smart device usage during the age of IoT was more convenient than ever, but it opened new windows of opportunity for cyberattacks. Legacy IPS has not been able to prevent dynamic, sophisticated threats from smart devices. This paper proposes an adaptive intrusion-prevention method for smart devices based on deep learning. Researchers introduce a hybrid framework that leverages a Convolutional Neural Network (CNN) to extract spatial features from network traffic time series and an LSTM network to handle sequential data at varying time steps, enabling the system to learn and adapt to changing attack trends. It trains and tests on the "Smart Home Intrusion Detection Dataset," a publicly available Kaggle data set comprising a sequence of common smart home network attack scenarios. It is developed using TensorFlow and PyTorch, trending deep learning frameworks, with Scikit-learn library support for data pre-processing, post-processing, and metrics. Our results confirm that the proposed model is unmatched in accuracy for intrusion prevention and detection compared with traditional machine learning models. The deep learning model's ability to learn and optimise makes it a potential candidate for enhancing the security of smart devices against advanced cyberattacks.
Keywords: Intrusion Prevention; Deep Learning; Smart Devices; Internet of Things (IoT); Cyber Attacks; IoT Environments; Detection Systems; Computational Efficacy; Sophisticated Threats.
Received: 12/01/2025, Revised: 03/05/2025, Accepted: 04/07/2025, Published: 12/12/2025
DOI: 10.64091/ATICS.2025.000214
AVE Trends in Intelligent Computing Systems, 2025 Vol. 2 No. 4 , Pages: 220-229