|
Author: Malinovsky Vadym Igorevich, PhD, associate professor in the Department of Information Protection, Vinnytsa National Technical University, Vinnytsa, Ukraine; Kupershtein Leonid Muckhailovich, PhD, associate professor in the Department of Information Protection, Vinnytsa National Technical University, Vinnytsa, Ukraine; Lukichov Vitaliy, PhD, associate professor in the Department of Information Protection, Vinnytsa National Technical University, Vinnytsa, Ukraine
|
Devices of the Internet of Things (IoT) in 2021-2022 years are widely cover all spheres of modern peoples life from household personal use devices to professional industrial systems and devices. The Internet of Things is a number of devices, such as personal mobile user devices (smartphones, bio trackers, fitness trackers, startwatches, others various devices for portable use) from individual personal use to number of industrial professional devices and nodes for IIoT (Industrial Internet of Things), are also exposed by modern cyber threats and data influences from Internet side and others communication sources.
Modern Internet of Things (IoT) devices and their architecture are include high integration intelligent ICs and multi level software, including IT high-tech decision and complex software modules. That provide a additional complex risks factors for their stability and normal functionality which are contain risk of functionality of IoT microcontrollers and communication paths and schemas with separate mobile operating systems or firmware in their composition [1] . The most of IoT devices problem – it’s their stability and reliability.
Modern IoT has a significant problem – has a complex information security [2] and data reliability risks for their functionality and stability of their information processes. With the growing popularity of smart devices and IoT services, the intensity of cyber threats is increasing. The trends of modern years indicate that the main threat factors in modern IoT devices are:
- complex cybesecurity of IoT devices and combined with them devices [1, 2] ;
- complex reliability risks of IoT devices and their modules;
- software and hardware core reliability of operating functionality, witch implements of the IoT platphorm functionality and IoT infrastructures functionality.
The latest modern technologies of Internet of Things devices (IoT, Internet of Things) and BYoD (Bring Your Owned Device) users' personal devices, such as smartphones, fitness bracelets, start watches, personal health monitoring devices, sensors, WiFi and Bluetooth headsets and quite widely cover all spheres of modern life from household personal use to industrial systems of professional specialized direction [2] .
The probability of failure-free operation of IoT or BYoD during time t is determined by some reliability function . The probability that the IoT or BYoD object will fail during time t characterizes the opposite property – unreliability and is expressed as: q(t)=1-p(t). Obviously, q(t) can be considered as a failure distribution function, its derivative:
It is the density of the distribution of uptime or density of failures.
The very function of reliability in IoT or BYoD is described by the main elements on which it depends:
where Pi (t)pi – single component of reliability for each of the factors affecting reliability (described in the list above); pi(∆t) – unit reliability component for each of the factors at the time interval ∆t= ti – ti-1 .
The meaning of this terminology can understanding if we proceed from the determination of reliability indicators by observation and statistical processing of failures of IoT or BYoD devices of a sufficiently large number of homogeneous objects. If you record the time during which each object worked before failure, you can determine the number of all those objects nt, the failure of which occurred over time ∆t. A fraction from division nt on the number of all investigated objects n gives an approximate value of the distribution function q(t)=nt/n, which is more accurate, the larger the number of objects n participated in the test. From here we can assume that if the system has n homogeneous elements, then the expected number of failures during time t will be equal to nq(t) and the expected frequency of failures is – nfr(t). Function fr(t) can be considered as the frequency of failures of homogeneous elements, related to their total number, which is expressed by the term "density of failures". Most often, the intensity of failures is taken as an indicator of the reliability of IoT or BYoD, which is equal to the ratio of the expected frequency of failures to the expected number of functional elements, i.e.
From here, the general law of reliability for IoT or BYoD can be expressed in terms of failure intensity:
The reliability of an IoT or BYoD system or device depends on its composition and structure, that is, on the number and quality of its constituent elements and interconnections. The source of unreliability is the failure of elements with the maximum function or failure density, which is part of the IoT or BYoD system.
Fig. 1 . Illustration of serial connections of data units in the software of IoT or BYoD
Illustration the simply data-flow model of the IoT or BYoD system, in which the failure of each element is happens independently and leads to the failure of the entire system (simple system), in the sense of reliability, a serial connection of elements is presented (Fig. 1). Although physically they can be connected as desired. According to the rule of multiplication of probabilities for independent events, the probability of fault-free operation of the system p(t) is equal to the product of probabilities pi trouble-free operation of it’s elements (i=1,2,...,n):
Thus, the stability indicators of the software product or software system, which consists of modular components, obtained as a result of the probability of error calculations – free stable operation and the average time of stable operation meet the necessary requirements.
Risk assessment of the emergence of cyber threats for information software modules in IoT are defined by sum of the complex unit probabilities of the occurrence of a cyber threat or unreliable functionalities. The criterion of a comprehensive assessment using the likelihood of the occurrence of a threat:
where ð³ pi – unit probabilities of the occurrence of a cyber threat for each of the main informational factors of threats; ki – correction coefficients for each threat factor; KS – complex correction coefficients for all threat factors.
It should be noted that for a comprehensive assessment of the cyber threats it will be fair 0 < P(λi)<1, and the higher the set of factors with the corresponding probabilities ri, the greater the total probability will approach 1. Accordingly, the conditions of stable work with the absence of risks are ensured by the evaluation of the indicator (stability coefficient):
The higher the indicator R`(R`ϵ 0 …1) – the better the conditions for the stability of the IoT software functionality and for IoT or BYoD information system in general.
References
[1] Temechu G. Zewdie, Anteneh Girma . IoT Security and the Role of AI/ML to Combat Emerging Cyber Threats in Cloud Computing Environment // Information Systems Journal. –21(4). –2020. – pp.253 – 263, DOI:10.48009/4.
[2] A. Girma. Analysis of Security Vulnerability and Analytics of Internet of Things (IoT) Platform // Information Technology – New Generations. – 2018. – vol. 738.– pp.101-104.
|
Ãîëîñóâàííÿ 
Öÿ ðîáîòà ë³öåíçóºòüñÿ â³äïîâ³äíî äî Creative Commons Attribution 4.0 International License
Çíàéøëè ïîìèëêó? Âèä³ë³òü ïîìèëêîâèé òåêñò ìèøêîþ ³ íàòèñí³òü Ctrl + Enter