Categorization of CoAP DoS Attack Based on One-Class Boundary Methods

Abstract

Denial of Service (DoS) attack over Internet of Things (IoT) is among the most prevalent cyber threat, their complex behavior makes very expensive the use of Datagram Transport Layer Security (DTLS) for securing purposes. DoS attack exploits specific protocol features, causing disruptions and remaining undetected by legitimate components. This paper introduces a set of one-class boundary methods such as Approximate Convex Hull, K-Nearest Neighborhood and One-Class Support Vector Machine for developing a categorization model in order to prevent IoT DoS attacks over the CoAP (Constrained Application Protocol) environments.