Modeling the Internet of Things Under Attack: A G-network Approach

• Published in: IEEE internet of things journal 2017-12, Vol.4 (6), p.1964-1977
• Main Authors: Sarigiannidis, Panagiotis, Karapistoli, Eirini, Economides, Anastasios A.
• Format: Article
• Language: English
• Subjects: Analytical models; Arrivals; Biological system modeling; Cybersecurity; Data acquisition; Data collection; Data integrity; Data loss; Data transmission; Digital media; G-networks; Internet of Things; Internet of Things (IoT); Jamming; Luminous intensity; Mathematical models; Model accuracy; modeling; Numerical models; Packets (communication); queuing theory; Routing; Security; Traffic information
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2017.2719623

This paper introduces a novel, analytic framework for modeling security attacks in Internet of Things (IoT) infrastructures. The devised model is quite generic, and as such, it could flexibly be adapted to various IoT architectures. Its flexibility lies in the underlying theory; it is based on a dynamic G-network, where the positive arrivals denote the data streams that originated from the various data collection networks (e.g., sensor networks), while the negative arrivals denote the security attacks that result in data losses (e.g., jamming attacks). In addition, we take into account the intensity of an attack by considering both light and heavy attacks. The light attack implies simple losses of traffic data, while the heavy attack causes massive data loss. The introduced model is solved subject to the arrival and departure rates in terms of: 1) average number of data packets in the application domain and 2) attack impact (loss rate). A comprehensive verification discussion accompanied by numerous numerical results verify the accuracy of the proposed model. Moreover, the assessment of the presented model highlights notable operation characteristics of the underlying IoT system under light and heavy attacks.