ANALYZING THE IMPACT OF MACHINE LEARNING-BASED SECURITY MEASURES ON QOS IN IOT NETWORKS

Abstract

 The emergence of the Internet of Things (IoT) and other innovative technologies has enabled the interconnection of devices worldwide, earning them the moniker "smart gadgets" due to their capabilities to send, receive, and process data. This technology is experiencing rapid growth, with a continually increasing user base. The success of IoT hinges on factors such as data transmission rates, quality of service (QoS) maintenance, and management of energy constraints in battery-operated devices. At the network level, QoS is evaluated based on metrics including end-to-end delay, throughput, jitter, and packet delivery ratio. With the proliferation of IoT devices, ensuring both device and data security in network communications becomes paramount. This paper delves into algorithms employed to safeguard the locations of source and sink nodes from potential breaches. Additionally, it investigates the impact of AODV protocols on the QoS offered by IoT networks. Malware poses significant threats in this context, prompting researchers, industry professionals, and end-users to seek effective countermeasures. Early and accurate prediction of malware behavior is crucial to mitigate potential damage. The research endeavors to combat malware using the K-Nearest Neighbors (KNN) algorithm, predicting its behavior and eliminating it. Utilizing these classifiers appropriately can significantly enhance prediction accuracy.