ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 8. Vol. 23. 2017

To the contents

V. I. Anikin, D. Sc. Professor, e-mail: anikin_vi@mail.ru, Volga Region State University of Service, O. V. Anikina, PhD, Assistant Professor, e-mail: blue-waterfall@yandex.ru,
Togliatti State University

The Visual Component Modeling of Fuzzy Inference Systems in Spreadsheets

It has been proved that any spreadsheet model is a graph — machine of some algorithm. The interpretation of the spreadsheet model as the algorithm graph-machine and the component modeling technology combined with the modern means of cells formatting do spreadsheets a unique environment for the visualization of algorithms structure, including fuzzy algorithms.
The library of components for the visual modeling of fuzzy logics without writing program code has been developed. The technology of component modeling of fuzzy logics has been tested on three fuzzy logics inference systems: Tipper (algorithms Mamdani, Larsen and Takagi—Sugeno), Pendulum (algorithms Mamdani and Takagi—Sugeno), Risk (algorithm Mamdani).
The presented results may find broad application in at least two areas: 1) for programmers developing new fuzzy logic algorithms and programs, 2) in laboratory practice on the disciplines of artificial intelligence to visualize the fuzzy logic theoretical bases, algorithms and systems.
Keywords: fuzzy logic, spreadsheets, component modeling, algorithm's graph-machine, algorithms visualization

P. 555–563

Î. À. Sosnina, Assistant Professor, e-mail: o.a.sosnina@mail.ru, A. D. Filinskikh, Assistant Professor, e-mail: alexfil@yandex.ru, A. S. Korotaeva, Master's Student, e-mail: korotaeva.as@gmail.com, Nizhny Novgorod State Technical University n.a. R. E. Alekseev

Comparing Approaches to Creating a Low Polygonal 3D Model

Considered in the article is the issue of creating low polygonal models, which are used in the development and animation of video games, when it is essential to save computer resources and when a detailed elaboration of the 3D object is not required. To address the issue of simulation performance and to ensure accuracy in modelling, the article studies diverse approaches to creating models and singles out their advantages and demerits. Proposed in the article is an approach that facilitates and expedites creating a low polygonal 3D model on the basis of an existing physical model. As exemplified by the creation of a fairy-tale character's model, the article makes a practical comparison of modelling with and without the use of auxiliary underlying planes carrying images. As a result, an effective way to create low polygonal 3D models is suggested.
Keywords: low polygonal model (low poly), 3D modelling, photogrammetry, 3D scanning, methods of modeling

P. 564–568

S. A. Inyutin, D. Sc., Full Professor, Moscow Aviation Institute (Nation Research University) (MAI), e-mail: inyutin_int@mail.ru

Method Calculation Characteristic of the Order Relation for Computer Modular Value

An iterative method for calculating the quantitative characteristic of the order relation for parallel computer data formats in modular reconfigurable computing systems is developed. The quantitative characteristic is intended for performance of separate operations in parallel computer arithmetic. The complexity of calculating the quantitative characteristic is a critical parameter for any software or technical implementations of modular arithmetic, which allows parallel and independent execution of ring algebraic operations on modular data formats on computational SIMD architectures. In the method, deductions for individual modules using diophantine relations are decomposed into regular finite sums, the terms of which are then element-wise added together with the verification of the criterion for completing the process of calculating the quantitative characteristic. The simulation of the method shows the convergence for 2—3 stages of summation. The iterative method allows under certain conditions to obtain a characteristic with a linear or logarithmic complexity of the algorithm for pyramidal summation on a parallel architecture.
Keywords: multiprocessor reconfigurable SIMD architecture systems, parallel computational process, linear and logarithmic computational complexity, parallel computer formats, numerical quantitative characteristic, regular finite sums

P. 569–574

D. V. Telpukhov, Ph. D., Head of Department of IPPM RAS, e-mail: e-mail: nofrost@inbox.ru,
R. A. Solovyev, Ph. D., Leading Researcher, Institute for Design Problems in Microelectronics, Zelenograd, 124681, Russian Federation

Development of a Method for Identical Fault Search in Logical Circuits

The problems of IC reliability are now becoming increasingly important. The idea of reliability is a complex integrated concept and includes structural, schematic and technological solutions, as well as solutions at the level of logic gates, register transmissions and microsystem as a whole.
This paper is devoted to preprocessing of logical circuits aimed at the reduction of computational costs of evaluating fault tolerance parameters, as well as the costs of circuit verification and testing. A method for identical fault search in logical circuits have been proposed, which makes it possible to estimate the effect of only one fault from the corresponding equivalence class of identical faults in subsequent verification and evaluation processes. In addition, these data can be used to construct self-test circuits.
This task was inspired by ICCAD 2016 Contest, within the frame of which software for identical fault search was written. The presented decision won the fourth place in the contest and was awarded honorable mention at the award ceremony. Note that in terms of correctness of equivalence classes' identification this solution was the best one presented, and was the only one that reached 100 % accuracy.
Keywords: equivalent circuits, ICCAD Contest, fault injection, identical fault search

P. 575–582

Yu. N. Fonin, Research Associate, e-mail: yurifonin@gmail.com, Moscow Institute of Physics and Technology (State University)

Using PPDL Language for Automatically Generation of the Tool Chain for Processors

The modern system on chip development technologies requires an iterative approach, where the hardware and software is modified several times to achieve the best characteristics for the specific benchmark. To accelerate the iterative development process a special high-level architecture description language PPDL has been developed. The paper describes the syntax of PPDL. A set of constructions for resource description and behavior of processor is present. Constructions for the declaration of registers, flags, memory buses, IO channels as well as groups of resources, called in PPDL as sets, are given. A behavioral part of the processor's description is present by instructions's description and pipeline description. Beside the PPDL constructions description, the paper describes experience of using PPDL for automatically generation of the tool chain components for NeuroMatrix processor. The example of using PPDL for the NeuroMatrix tool chain development showed that PPDL allows to accelerate a tool chain development about three times compare to manual development of the separated components.
Keywords: PPDL, architecture of microprocessor, tool chain, architecture description language

P. 583–588

R. E. Asratian, Leading Researcher, e-mail: rea@ipu.ru, V. N. Lebedev, Head of Laboratory, e-mail: lebvini@ipu.ru V. A. Trapeznikov Institute of control sciences of Russian academy of sciences

Internet Service for Pipeline Processing of Protected Information Queries in Multi-Network Environment

The principles of the organization of the new network service — Protected Message Service (PMS) — intended for protected queries processing in the distributed information systems oriented on operation in complex multi-server and multi-network structures are considered. Distinctive features of service are tight integration of functions of information data protection with functions of in­formation exchange and information requests routing for multi-server processing. From the client point of view the service architecture is based on two main program classes: "Protected message" and "Network Connection". This classes offer necessary functionality not only for creating and protecting messages, but also for transferring them to remote server via established network connections for processing. Contrary to web-cervices, based on remote function call model, PMS-service is based on message processing model: all service functions receive object of "Protected message" class as a parameter and return another object of the same class as a result of processing. This approach opens a basic possibility of the organization of pipeline processing of one protected message by the sequence of service functions in a manner when the output of each function is forwarded to the input of the following one, and the result of last function returns to the client. The principles of the organization of pipeline processing of queries in a chain of servers are considered. The emphasis on routing of queries in the complex multi-network structures including a set of private local area networks is especially placed. Some experimental estimates of speed of pipeline multi-server processing are presented.
Keywords: distributed systems, Web-technologies, Internet-technologies, network interactions, data security

P. 589–597 

A. V. Kugaevskikh, PhD, Associate Professor, e-mail: a-kugaevskikh@yandex.ru Tyumen State Medical University, Tyumen, Russia, 625000

Comparison Gabor Filter Parameters for Efficient Edge Detection

In this paper I proposed generalized formula of Gabor filter. It was made according to the extensive review of sources. I carried out comparing of different coefficients influencing the form of a kernel on a DC component metrics. Thus the optimum formula was defined. The significant parameters (wavelength and scale of the filter) influencing quality of edges detection were revealed. The metrics of assessment on which comparing of different relations of significant parameters is carried out was offered. As a result the optimum ratio of significant parameters was found. It is proved that the equation offered by Petkov is optimum (equation 8). That was confirmed by the calculation of the DC component and the analytical comparison of the influence of different wavelengths on the form of the kernel graphs. Also, the description of the selection methods for the ratio of the wavelength and filter scale was performed. Quality assessment of edge detection metrics SNRe was proposed. Its application made possible to speak of the optimality ratio the use of the half-response spatial frequency bandwidth (equation 23).
Keywords: image analysis, gabor filter, wavelength, filter scale, edge detection, dc component

P. 598– 605

S. M. Avdoshin, PhD, Professor, Head of Software Engineering School, Faculty of Computer Science, HSE, e-mail: savdoshin@hse.ru, E. Yu. Pesotskaya, Docent, Faculty of Computer Science, HSE, e-mail: epesotskaya@hse.ru National Research University Higher School of Economics (HSE)

Internet of Things: Retail

The paper describes the possibility of using the Internet of Things in retail industry as well as potential of using internet of things in Russia. The research objective is to identify the possibilities of using Internet of Things in retail process such as supply, logistics, client service and marketing and to give the examples for getting economy and cost reduction in industry. The author provides an overview of possible tools and insights that transform retailers' businesses and recommendations of using internet of things technology in retail. Authors analyze IoT possibilities that can enhance and refine business processes, decrease operational expenditures, better understand and communicate with consumers, taking into considerations existing aspects and current worldwide trends of global transformation, digitalization and cloud technologies.
Keywords: Digital Transformation, Internet of Things, Automation, Cloud Technologies, Mobile, Mobile application (apps)

P. 606–612

D. A. Pechnikov, Associate Professor, e-mail: 19pda72@bk.ru Naval Academy named after Admiral of the Fleet of the Soviet Union NG Kuznetsov, St. Petersburg

Using the Methods of the Interval Estimation of the Results of the Control of Knowledge in Computer Testing Systems

This article discusses the use of models and procedures for interval estimation of the results of control of knowledge in computer systems using methods based on the laws of binomial distribution and the geometric distribution as special case of the negative binomial distribution. The proposed approach to the transition from the empirical point estimates of the relative frequency of a suc­cessful execution of the test tasks to the interval estimates of the probability of this success to determine the margin of error and confidence interval of the estimates. The developed model criteria-oriented assessment of results of knowledge control, which is focused on implementation in computer systems testing.
Keywords: Interval estimation, confidence intervals, binomial distribution

P. 613– 618

L. M. Kadyeva, Student, Mathematician, e-mail: leylakadyeva@gmail.com, S. E. Levin, Chief Technical Officer, PKCC-PS, e-mail: levin@pkcc-ps.ru, Yu. N. Okrent, Chief Scientific Officer, e-mail: yakov.okrent@pkcc-ps.ru, Russian Corporation of Communication — Software Systems, Moscow, 105005
D. I. Loskutov, Student (MAI), e-mail: dmitry.loskutov@pkcc-ps.ru, A. M. Romanenkov, PhD, Associate Professor, e-mail: romanaleks@gmail.com, Moscow Aviation Institute (National Research University), Moscow, 125080

The Development of Neuro-Fuzzy Module Calculating the Priority Dispatching Tasks to Railway Station

The theoretical position, methodological procedures and recommendations aimed at identifying fuzzy identification system parameters. There is analysis of the collected statistical data intended for training of a neural network. The considered model is tested for adequacy (R2 criterion). The structure of a neural network is shown. Spend the development of neuro-fuzzy module for calculating the dispatch task priorities in the decision support system of dispatching service railway station with a complex topology of ways. An example of the program module and analysis of the results are shown.
Keywords: identification of nonlinear dependencies, neuro-fuzzy network, Sugeno algorithm of fuzzy systems design, priority queue, the problem of finding the optimal sequence of operations

P. 618–624

To the contents