RU
UA
EN

Abstract

When writing this essay master's work is not yet completed. Final Completion: June 2019. Full text of the work and materials on the topic can be obtained from the author or from his scientific adviser after the specified date.

Content

Introduction

In the development of the industry of electronic devices, there is a tendency aimed at achieving the highest possible compactness and mobility of devices by reducing the size and weight of the modules and materials used. This trend is accompanied by a trend of increasing their intelligence. Such devices are based on embedded systems that implement the functionality of the device and operate under the control of special software. An important requirement for such systems is the ability to process incoming events at fixed intervals, i.e. execution of software within certain real-time boundaries. In the quest for compactness, the computing power of such systems is reduced. The established restrictions on the program execution time and the maximum possible frequency of the microprocessor determine the maximum complexity of the algorithm. In the case of restrictions on the size and weight of the device, as well as the output of the algorithm beyond the maximum allowable labor intensity, the physical implementation of the device is in doubt.

One of the approaches to solve this problem can be the removal of labor-intensive parts of the algorithm in the form of functions to external computing resources, with subsequent access to them. The use of software distribution technologies in device design implies the need to use a data interface to access a remote piece of software.

At the same time, the type of available interfaces and the type of possible resources used to execute the remote part of the software introduce some uncertainty that prevents the exact time required for the remote part of the software to be executed relative to the embedded system. The uncertainty of these values when considering the interaction process of distributed embedded systems increases accordingly.

1. Theme urgency

A new trend in the development of electronic devices is the emergence of smart things that have the functions of interaction with each other.

The development of the industry of autonomous mobile devices, such as: unmanned aircraft, ships and cars, user robots and quadcopters. These devices include an embedded control system, a set of sensors and actuators, which together form a cyber-physical system. Today, cyber-physical systems are being actively implemented in all spheres of human activity, while ensuring a positive economic effect [1].

2. Goal and tasks of the research

The presence of a mathematical model allows us to determine the possible time spent on performing a particular function using remote software in the process of interaction of embedded systems, relative to a specific instance of the system, during operation. Its use allows you to more smoothly schedule real-time tasks.

This master's work is devoted to the development of a model for the interaction of embedded systems with distributed resources based on mathematical and physical models.

It is planned to consider the methods of using remote software and evaluate the results obtained.

3. Method for determining the efficiency of using cloud functions based on a mathematical model

The complexity of the algorithm is determined by the required number of processor operations and I / O operations [12]. Suppose that in the real-time system in question, the processing speed of commands is equal to the frequency of the base microprocessor, and the processing time for one command is 1 clock cycle. Then, if the system has a limit on the reaction time and at a constant frequency, you can determine the maximum complexity of the algorithm using the following formula:

(1)

где
– maximum complexity of the algorithm,
– command processing frequency,
– reaction time to an external event.

To achieve efficient allocation of processor time among tasks, when using a distributed algorithm, the total value of the data transfer time (when passing the function argument), the actual execution time of the function and the time to get the result should not exceed the execution time of the same function by the microprocessor. Therefore, the condition:

(2)

где
– cloud data transfer time,
– cloud function execution time,
– time to get the result from the cloud function,
– microprocessor processing time,
– the number of bytes transmitted,
– the number of bytes received,
– complexity of the algorithm.

To verify the fulfillment of condition (2), it is required to determine the complexity of the algorithm for which the calculation is performed, and to determine the forms of time functions.

It is worth noting that the theoretical calculation of the complexity of the algorithm may have a large error due to the fact that in such a calculation it is necessary to use a probabilistic analysis of the transition with the branching of the algorithm. At the same time, the complexity of the algorithm can be empirically determined by determining the execution time of the algorithm by the microprocessor of the real-time system.

Time functions, in turn, depend on the data interface, protocol and communication media.

Conclusion

The development of technology has reached the point at which the set of tasks to be solved is shifted to the device, while minimizing human participation. In the process of developing the intellectual component of devices, the requirements for computing capabilities of embedded systems are increasing more and more. Due to the lack of a single focal point, the coordination issues of the devices fall on their algorithms.

To date, there are already technologies that allow you to expand the computing capabilities of embedded systems and minimize their size, thereby achieving the possibility of implementing such devices. However, issues related to the implementation of algorithms distributed in this way, within the specified time limits, are still open. Thus, it is necessary to develop a model of interaction among themselves of embedded systems with distributed resources, analysis and study of its responses for the successful formation of algorithms for the interaction of distributed embedded systems.

References

  1. Куприяновский В.П. Кибер-физические системы как основа цифровой экономики / В.П. Куприяновский, Д.Е. Намиот, С.А. Синягов // International Journal of Open Information Technologies ISSN: 2307-8162 vol. 4, no. 2, 2016.
  2. 2016. Число гипермасштабируемых ЦОДов в мире достигло трехсот [Электронный ресурс]. — Режим доступа: http://www.tadviser.ru/index.php/Статья:ЦОД_(мировой_рынок)_Коммерческие_дата-центры
  3. G. Coulouris, J. Dellimore & T. Kindberg, Distributed Systems — Clouds, Edition 2, 1994.
  4. Что такое облачные вычисления? Руководство для начинающих [Электронный ресурс]. — Режим доступа:https://azure.microsoft.com/ru-ru/overview/what-is-cloud-computing/
  5. Mell, Peter and Grance, Timothy. The NIST Definition of Cloud Computing. Recommendations of the National Institute of Standards and Technology. NIST (20 October 2011) [Электронный ресурс]. — Режим доступа: https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-145.pdf
  6. Соломаха С.С., Мальчева Р.В., Дегтярева И.И. — Применение облачных вычислений в системах реального времени // Материалы IX Международной научно-технической конференции Информатика, управляющие системы, математическое и компьютерное моделирование (ИУСМКМ-2018). — Донецк: ДонНТУ, 2018. — С. 182-186; [Электронный ресурс]. — Режим доступа: http://iuskm.donntu.ru/electronic/iusmkm2018.pdf
  7. Fog Computing and the Internet of Things: Extend the Cloud to Where the Things Are // Cisco White Paper, 2015. — 6 c. [Электронный ресурс]. — Режим доступа: https://www.cisco.com/c/dam/en_us/solutions/trendsiot/ docs/computing-overview.pdf
  8. Туманные вычисления (Fog Computing), как составная часть 5G // Телеком и ИТ. [Электронный ресурс]. — Режим доступа: https://shalaginov.com/2018/07/24/4510/
  9. FPGAs Accelerating IoT Gateway and Infrastructure Tiers [Электронный ресурс]. — Режим доступа: https://www.aldec.com/en/company/blog/125--fpgas-accelerating-iot-gateway-and-infrastructure-tiers
  10. Fog Computing for the Internet of Everything, Mentor, July 28, 2017 [Электронный ресурс]. — Режим доступа: http://www.rtcmagazine.com/fog-computing-for-the-internet-of-everything/
  11. Жирков А. Интернет вещей и облачные технологии Eurotech // Современные технологии автоматизации. — 2015. — № 2. — С. 6-12. https://tp.prosoft.ru:443/docs/shared/Техпортал/Eurotech/Техническая%20статья/Интернет%20вещей%20и%20облачные%20технологии%20Eurotech.pdf
  12. Оценка трудоемкости алгоритма [Электронный ресурс]. — Режим доступа: http://life-prog.ru/1_56130_otsenka-trudoemkosti-algoritma.html