GuardiaNet: Intelligent Device Identification in Smart Home Networks
Loading...
Date
2024
Journal Title
Journal ISSN
Volume Title
Publisher
UMT, Lahore
Abstract
We live in the world of smart devices and the protection and safety of our networks are vital in the present world. This study explores the process and procedure of developing and using an intelligent system known as GuardiaNet for detecting probable threats from IoT devices in the network environment. Thus, the foundation for GuardiaNet is derived from a rigorous and detailed process of feature engineering that would seek to capture almost any form of unconventional and abnormal behavior of communication that an IoT device might produce. Techniques of selection, transformation, and creation of features at GuardiaNet enable the utilization of detailed features of the network traffic for identifying patterns that reflect device identity. Based on this, GuardiaNet uses the following machine learning techniques: AdaBoost, Decision Trees, Neural Networks, and Linear Regression necessary for the real-time IoT device classification and identification. These algorithms not only help in identifying the devices accurately but also let GuardiaNet to grow according the changing environment and nature of IoT networks. One of the significant aspects to consider in the framework of presented research is the implementation of user-friendly and practical solutions. The interfaces of GuardiaNet are straight forward and fully integrated so no matter how novice the network administrator is, he or she will not have a problem in initially implementing the system or in operating it. In addition, GuardiaNet is highly portable and comes in options, and it can readily be adjusted to thus ensuring that it meets the requirements of the organization. In the design of GuardiaNet, confidentiality is an important factor of consideration. The best practices coupled with anonymization mechanisms are implemented in order to protect data privacy and respect users’ rights, observing legislative trends and requirements. The performance of GuardiaNet is tested using real datasets including, CICIoT 2022 dataset to establish its efficiency in different IoT contexts. Subsequently, this paper proves how, through concrete performance metrics and validation process, GuardiaNet can efficiently detect, categorize and neutralize possible security threats from IoT devices.