Internet of Things (IoT) Cybersecurity Enhancement through Artificial Intelligence: A Study on Intrusion Detection Systems

Gangadhar Sadaram, Manikanth Sakuru, Laxmana Murthy Karaka, Mohit Surender Reddy, Varun Bodepudi, Suneel Babu Boppana, Srinivasa Rao Maka

Citation: Gangadhar Sadaram, Manikanth Sakuru, Laxmana Murthy Karaka, Mohit Surender Reddy, Varun Bodepudi, Suneel Babu Boppana, Srinivasa Rao Maka, "Internet of Things (IoT) Cybersecurity Enhancement through Artificial Intelligence: A Study on Intrusion Detection Systems", Universal Library of Engineering Technology, Special Issue.

Copyright: This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Computers networks are very vulnerable and can be threatened by hackers, viruses and other immoral entities. Intrusion detection is a crucial component of network security since it is an active defence technique. limited accuracy, limited detection effectiveness, high false positive rate, and incapacity to deal with new forms of intrusions are some of the challenges faced by traditional intrusion detection methods. These problems may be solved by suggesting a real-time network intrusion detection system that uses ML. With the use of the CICIDS2017dataset, which contains both benign traffic and a range of attack types, this research introduces an Intrusion Detection System (IDS) intended to improve IoT cybersecurity. For the selection of the features, which are important and are not numerous, PCA is applied. The SMOTE is a tool for addressing class imbalance. Among these models, DenseNet, KNN, SVM and DeepGFL are some of the models whose performances are measured compared with a set of performance metrics such as recall, accuracy, precision and F1-score. Therefore, DenseNet provides the highest level of efficiency and reliability for all of the research and development models proposed, with a perfect back-and-forth recall of 98.2%, accuracy of 99.12%, and precision of 98.6%. Therefore, the results confirm DenseNet’s applicability in detecting intrusions in IoT networks. In subsequent research, more attack types will be integrated into the dataset, and real-time integration will be considered, as well as using deep reinforcement learning for dynamism for threat detection at the IoT system level.


Keywords: Internet of Things (IoT), Cybersecurity, Network Security, Intrusion Detection System (IDS), Attack, Machine learning.

Download doi https://doi.org/10.70315/uloap.ulete.2022.001