Aidynov T., Altaibek M., Tleuberdin S., Nurusheva A., Satybaldina D., Abisheva G.
Application of Deep Learning Techniques for Automatic Detection of Network Security Threats in Internet of Things Environments // Proceedings of the 2025 IEEE European Technology and Engineering Management Summit (E-TEMS). — 2025. — Conference held 26–28 May 2025. — DOI: 10.1109/E-TEMS64751.2025.11239312.

Abstract: This paper investigates the application of machine learning and deep learning techniques for strengthening cybersecurity in Internet of Things (IoT) environments, focusing on the IoT-23 dataset and the CRAFTED (Cooja RPL Attack Framework Test and Evaluation Dataset). Multiple models are developed to classify and predict a wide range of IoT attack types and parameters. Experimental evaluation includes deep learning architectures such as Long Short-Term Memory (LSTM) networks and Residual Networks (ResNet), compared against traditional machine learning models. Results show that deep learning methods are effective in capturing complex patterns in multidimensional data, while classical models such as Random Forest and Decision Trees provide better efficiency in terms of processing speed and real-time classification performance. Feature selection and hyperparameter tuning further improve model accuracy. The study highlights the trade-off between accuracy and computational efficiency, providing insights for designing practical IoT cybersecurity systems.

Link / DOI: https://doi.org/10.1109/E-TEMS64751.2025.11239312