SUBSTANTIATION OF THE PRELIMINARY SELECTION OF ARCHITECTURE OF DATA PROCESSING SYSTEM USING FUZZY LOGIC

Abstract

The purpose of the work is to formulate an approach to the preliminary justification for choosing the type of architecture of the data processing and control system. System architecture is a way of constructing and organizing its functioning in the execution of data processing and control programs. The quality of the architecture can be viewed from the standpoint of accepted efficiency criteria such as, for example, productivity, volume of resources, cost of processing and others. Initial data for making decisions on the choice of the best architecture are the characteristics of these problems, processing algorithms, characteristics of acceptable types of architecture of computing devices, conditions and requirements for the organization of computing processes and control processes, processing procedures, their characteristics and parameters, features of the software environment, development tools and modification of software solutions. The uncertainty caused by the future aspects of the data processing system’s functioning and conditions of use, as well as constantly changing external and internal factors, necessitates the use of approaches to design the data processing architecture from the standpoint of reducing the risk of making unreasonable decisions. Therefore, there is a need for data processing as part of a workload that changes over time, manifesting itself in the totality of data processing tasks and their intput, and in the necessary processing procedures. These conditions form a data processing environment for which a processing system with an adequate architecture can be used. The degree of adequacy of the architecture of such a system can be estimated from the standpoint of the selected criteria and the degree of their agreement. The system architecture options that match the agreed solutions are a subset that provides sound decision choices that can be made with efficiency evaluations. Given the growing interest of customers in the development of cloud-based computing systems, the justification and choice of data processing system architecture using cloud-computing services is of particular relevance. It may take a few minutes to prepare such systems for application. Therefore, to improve the quality of justification for the previous choice of architecture of the data processing system, it is proposed to use the procedures of the fuzzy logic. An example of numerical calculations and an analysis of the results obtained are offered to illustrate the approach.