An Algorithm based on Deep Learning for Intrusion Detection in IoT

Zoya Firdouse, K. Jayasree

Abstract: In order to improve intrusion detection system (IDS) performance in Internet of Things (IoT) environments, an innovative framework is presented in this study. By leveraging the ToN-IoT telemetry dataset, which includes environmental sensor data such as temperature, pressure, and humidity, the framework aims to improve the classification and prediction of cyber attacks. This approach integrates machine learning and deep learning algorithms, including Random Forest (RF), Decision Tree (DT), k-Nearest Neighbors (KNN), Gradient Boosting, and Long Short-Term Memory (LSTM), to provide a comprehensive analysis. By combining telemetry data with traditional network data, the proposed framework offers a more holistic view of the system, thereby enhancing the accuracy of intrusion detection. This method helps in reducing false positives and improving contextual awareness, allowing for the differentiation between legitimate environmental changes and potential cyber threats. The results of our study emphasize the significance of combining a variety of data sources and cutting-edge algorithms to strengthen IoT systems against cyber threats by showing how the ToN-IoT telemetry dataset greatly enhances the detection and prediction of Cyberattacks.

Keywords: Random forest (RF), Decision Tree (DT), Internet of Things, IoT, intrusion detection systems, KNN, Gradient boost, LSTM

How to Cite?: Zoya Firdouse, K. Jayasree, "An Algorithm based on Deep Learning for Intrusion Detection in IoT", Volume 13 Issue 8, August 2024, International Journal of Science and Research (IJSR), Pages: 1073-1078, https://www.ijsr.net/getabstract.php?paperid=SR24813173848, DOI: https://dx.doi.org/10.21275/SR24813173848