The Machine Learning Algorithms for DDoS Attack Detection on IoT Network Layer

Abstract

Network communication and infrastructure security now faces additional difficulties because of the widespread use of Internet of Things (IoT) devices. Distributed Denial of Service (DDoS) assaults are one of the many risks that Internet of Things (IoT) ecosystems must contend with. They can seriously jeopardize the availability and dependability of these networked devices. This study employs a machine learning algorithms to detect DDoS attacks at the IoT network layer. In this study, author employ Linear Regression, Random Forest and Decision Tree algorithms—a type of machine learning—to present a unique strategy for detecting and thwarting DDoS attacks. Considering crucial elements like packet counts, packet sizes, source and destination IP addresses, protocol types, and IoT- specific qualities, a detection threshold is established using the Hybrid Algorithm. An alert is raised, and protective procedures are started for the IoT ecosystem when this threshold is crossed. Our findings highlight the necessity of ongoing model updates and monitoring to respond to changing DDoS attack strategies.