Classification of Denial-of-Service Attacks in IoT Networks Using AlexNet

Abstract

The Internet of Things (IoT) plays an essential role in managing our work from home to aircraft. It is the functionality to network electronics in a standard manner wherein the linked devices transfer information into the cloud over the Internet with TCP/IP. The Internet of Things (IoT) is a powerful tool for building adaptable and cost-effective systems. Still, it has various security risks, including denial-of-service (DoS) and distributed denial-of-service (DDoS). Although many security elucidations have been implemented for networked devices, an effective apprehension system for IoT networks is still needed. Incorporating deep learning-based techniques into the detection systems will make them reliable and effective. Our detection system utilized the potential of convolution neural networks (CNN) in image processing. First, a public NetFlow dataset was converted to an image dataset and trained into an AlexNet CNN model. The proposed method can identify the attacking patterns of DoS and DDoS with an accuracy of 94%, which is 7% more than that of the ResNet CNN model.