Analytical model for availability assessment of IoT service data transmission subsystem

Abstract

This paper discusses the role of safety, availability, and dependability of Internet of Things (IoT) data services dedicated to the monitoring and control of objects in a physical world. These services are designed to be available to devices and users on request at any time and at any location. The Internet of Things differs from today’s global Internet in a number of ways. For instance, the networks are typically unmanaged, most of IoT services are safety critical, the link layers are optimized for low power usage, and most nodes have to be implementable in a lightweight manner. The design of IoT data services should take into account that this type of system has huge scalability (hundreds, thousands of clients). On the one hand this leads to significant financial costs for maintenance the IoT infrastructure, on the other hand, is necessary to provide the appropriated availability and safety level. In addition, today most of IoT services are operated in a critical system such as oil and gas industry, smart cities, medicine, the financial sector. According to this point an actual question of availability and safety estimations IoT services appears, because the fault of IoT system part can lead to huge financial losses or to observed objects death. In this paper, the model of IoT services as queue network is proposed that allows estimating the availability and safety of these systems.