Boosting-based DDOS Detection in Internet of Things System

Abstract

The Internet of Things (IoT) is an innovative communication framework due to its wide variety of applications. Existing systems, such as home networks made up of various IoT devices, still struggle to resist Distributed Denial-of-Service (DDoS) attacks. The IoT network is one smart environment that this research helps to make more secure. IoT offers a vast network to connect objects globally in an effort to lessen human effort & make a digitalized environment easier to operate. DDoS attacks, which recently have targeted many IoT networks & resulted in large losses, are one of the main risks that can completely disable a targeted system's availability. Two machine learning techniques, namely the LGBM & extreme gradient boosting, were used to identify the attacks that are currently taking place in the dataset (XG Boost). These techniques were chosen due to their superior capacity to forecast in large data volumes in a shorter amount of time than other techniques. The accuracies attained by LGBM & XG Boost in 30 & 229 seconds(s), correspondingly, was 94.88% & 94.89%.