What is it about?
This paper proposes a hybrid framework for classifying attacks and anomalies in an IoT network using both machine learning and deep learning techniques. The framework includes a pre-processing stage for data balancing and feature selection, and a classification stage using both machine learning and deep learning models. The use of a hybrid approach to combine the results of both models improves the overall performance of the intrusion detection system.
Featured Image
Photo by Markus Spiske on Unsplash
Why is it important?
The scale of devices connected to an IoT network is increasing enormously. Devices can communicate with each other over wired and wireless channels in a synchronous or asynchronous fashion. These communication channels can pose security risks to IoT devices and can introduce novel vulnerabilities and challenges. It is important to classify attacks and anomalies in an IoT network to ensure the security and privacy of the network. IoT networks can be vulnerable to various types of attacks and anomalies, which can disrupt the operation of the network and potentially compromise the security of the devices connected to it. By using machine learning and deep learning techniques to classify attacks and anomalies, it is possible to detect and prevent such incidents, helping to ensure the smooth and secure operation of the IoT network. Additionally, the use of machine learning and deep learning techniques can enable the system to adapt and learn to recognize new types of attacks and anomalies, making it more effective at detecting and preventing such incidents over time.
Perspectives
The proposed framework for classifying attacks and anomalies in an IoT network using machine learning and deep learning techniques could potentially be a useful approach for improving the security and reliability of such networks. The use of both machine learning and deep learning techniques can enable the system to handle both binary classification tasks (distinguishing between normal and anomaly data) as well as multi-class classification tasks (distinguishing between different types of attacks). Additionally, the use of techniques such as SMOTE and PSO for addressing data imbalance and improving feature selection can help to improve the accuracy and efficiency of the system. Overall, the proposed framework seems like a promising approach for addressing the security challenges faced by IoT networks.
Naser Ezzati-Jivan
Brock University
Read the Original
This page is a summary of: Deep learning driven anomaly based intrusion detection system for IoT, November 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3565386.3565493.
You can read the full text:
Contributors
The following have contributed to this page
