Internet of Things Security: An Experimental Analysis of Detecting Denial of Service Attacks with Machine Learning
Mohd Zain Khan, Mohammad Ubaidullah Bokhari, Takreem Fatima Khan · 2024
The Internet of Things (IoT) is one of the technologies with a high growth rate in the world and has a profound effect on the lives of people in many different ways. The growing number of IoT devices has created new challenges in securing connected systems, as the number of IoT devices increases the potential of malicious attacks targeting these devices and the network they are part of. One of the malicious attack known as Denial of Service (DoS) attack is an intentional effort to interfere with regular operations of IoT devices by flooding the system with excessive traffic or by taking advantage of security gaps. The goal of this type of attack is to prevent authorized users from accessing network infrastructure or IoT devices. Because of the malicious attacks associated with the Internet of Things, researchers and business leaders have recently expressed a strong interest in detecting possible intrusions in IoT networks. Machine learning plays an essential role in detecting IoT attacks by leveraging its capacity to analyze complex patterns and behaviour within the vast and dynamic datasets generated by interconnected devices. This paper suggests the experiment to detect the DoS attack using machine learning approach using CICIDS2017 dataset. The machine learning classifiers like RF, SVM, XGBOOST, NB are used to train the model in which RF perform the best machine learning classifier having 99.97% of accuracy.