350 rub
Journal Information-measuring and Control Systems №3 for 2010 г.
Article in number:
The period problems planning while distributed information processing
Authors:
N. A. Alexeev, O. V. Bogdanova, I. V. Kovalev, R. Yu. Tsarev
Abstract:
The modern approach to creation of the distributed programmno-information technologies for corporate and industrial structures consists in association in uniform system or a network of set of processing means of storage, information and management processing. At a stage system-technical designing of one of the important problems the problem of formation of algorithms of the distributed processing and management is. On the set structure of hardware-software means it is necessary to carry out a choice of system and applied programs, structures of data and ways of interaction of these components providing the set mode of application of programmno-information technologies. Thus it is necessary to consider, that in the distributed systems the mode of real time assumes limitation of time of the answer of a control system on inquiry of object. It was supposed, that the processor is the unique resource necessary for performance of works. At testing of the given methods, problems were divided into two classes. In the first-frequencies of problems are multiple more than to two base frequencies, and in the second - no more than to two base frequencies. Reduction of frequencies of problems or times of their performance can lead to increase in number of demanded processors. And on the contrary, the demanded number of processors can be reduced at the expense of increase in frequencies of problems or performance times, i.e. by increase in congestion of the processor. The considered algorithms can be realised both by means of models of one class, and with use of the multicomponent network model including determined and stochastic components. The concept of multicomponent network model with central logic allows to unite various program components of modelling-algorithmic support with a uniform database and to provide effective formation of algorithms of the distributed processing and management taking into account periodicity of problems.
Pages: 11-14
References
  1. Барский А. Б. Параллельные процессы в вычислительных системах. М.: Радио и связь. 1990.
  2. Воеводин В. В. Математические модели и методы в параллельных процессах. М.: Наука. 1986.
  3. Ковалев И. В., Царев Р. Ю. Моделирование и оптимизация параллельных процессов в информационно-управляющих системах: Учебное пособие. Красноярск: ИПЦ КГТУ. 2003.
  4. Слепцов А. И., Юрасов А. А. Автоматизация проектирования управляющих систем. Киев: Техника. 1986.
  5. Олифер В. Г., Олифер Н. А. Компьютерные сети. Принципы, технологии, протоколы. СПб.: Питер. 1999.
  6. Олифер В. Г., Олифер Н. А. Сетевые операционные системы. СПб.: Питер. 2001.
  7. Филлипс Д., Гарсия-Диас. А. Методы анализа сетей: Пер. с англ. М.: Мир. 1984.
  8. Gonzales, J. M., Deterministic Processor Scheduling // Computing Surveys. 1977. V. 9. No. 3.
  9. Neumann, K., Stochastic Project Networks. Temporal Analysis, Scheduling and Cost Minimization // Lecture Notes in Economics and Mathematical Systems. Springer-Verlag. 1990. No. 34.
  10. Phillips, D. T., Garsia-Diaz, A., Fundamentals of network analysis. Prentice-Hall, Inc. Englewood Cliffs. New Jersy. 1981.