«INFORMATICS AND APPLICATIONS»
Scientific journal
Volume 12, Issue 4, 2018

Content | About  Authors

Abstract and Keywords.

MODELING MOVEMENT OF DEVICES IN A WIRELESS NETWORK BY RANDOM WALK MODELS

• K. E. Samouylov  Peoples' Friendship University of Russia (RUDN University), 6 Miklukho-Maklaya Str., Moscow 117198, Russian Federation, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• Yu. V. Gaidamaka   Peoples' Friendship University of Russia (RUDN University), 6 Miklukho-Maklaya Str., Moscow 117198, Russian Federation, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• S. Ya. Shorgin  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The authors overview mobility models which are applicable for simulation of movement of users' devices in a fifth generation (5G) wireless network. Mobility patterns that are typical for simulating the behavior of users of a wireless ad hoc network are considered. The features of the models are discussed, both for individual motion of objects and for motion of groups of objects, from the point of view of appliance to the analysis of interference in wireless networks. The purpose of the paper is to propose a number of mobility models for making an informed decision when choosing a model for evaluating the quality of service in 5G wireless networks. The authors present simulation results that illustrate the method of estimation of the key performance quality parameter, i. e., signal to interference ratio. For illustration, the developed complex of analytical and simulation models is used for simulation of movement of shopping moll customers with the help of the grid random walk mobility model.

Keywords: entity mobility model; group mobility model; ad hoc network simulation; signal to interference and noise ratio; SINR

Comparison of two active queue management schemes through the M/D/1/N queue

• M. G. Konovalov  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• R. V. Razumchik  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Peoples' Friendship University of Russia (RUDN University), 6 Miklukho-Maklaya Str., Moscow 117198, Russian Federation

Abstract: The paper focuses on giving the first in the literature numerical evidence that the stationary performance characteristics of single-server queues with the general renovation mechanism may be as good as of single-server queues with the RED-type active queue management mechanisms (AQM). Comparison is made in the queueing theory context: the basic model is the M/D/l/N queue. The characteristics reported are: the loss ratio, average system size, and average number of consecutive losses along with the standard deviations. Numerical results are based on the well-known facts and some new analytic results, presented in the paper.

Keywords: queueing system; active queue management; RED; renovation

DETERMINING THE EXTREMES OF PRECIPITATION VOLUMES BASED ON THE MODIFIED "PEAKS OVER THRESHOLD" METHOD

• A. K. Gorshenin   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, Leninskie Gory, Moscow 119991, GSP-1, Russian Federation
• V. Yu. Korolev   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, Leninskie Gory, Moscow 119991, GSP-1, Russian Federation

Abstract: The problem of correct determination of extreme observations is very important when studying meteorological phenomena. The paper proposes ascending and descending methods for finding the threshold for extremes based on the Renyi theorem for thinning flows and the Pikands - Balkema - De Haan results. Using the observation data for 60 years for Potsdam and Elista, it is demonstrated that the ascending method can present excellent results for daily precipitation but for volumes of wet periods, the descending method should be used. The results of such nonparametric approaches are compared with the parametric criterion proposed in the previous papers by the authors.

Keywords: precipitation; wet periods; extreme values; thresholds; Renyi theorem; Pickands - Balkema - de Haan theorem; testing statistical hypotheses; data analysis

OPTIMIZATION OF BUFFER MEMORY SIZE OF SWITCHING NODE IN MODE OF FULL MEMORY SHARING

• Yа. M. Agalarov   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The problem of optimizing the volume of the buffer memory of the network switching node with the CP (Complete Partitioning) is considered, which uses a scheme for the complete memory sharing between the output transmission lines, taking into account packet loss due to memory overflow and average packet delays. The properties of the solution of the problem are obtained, and an algorithm for searching for the optimal volume is developed for the switching node model with Poisson incoming flows, common service time distributions, and single-channel output transmission lines. Within the given model, an algorithm for finding the optimal buffer memory allocation plan for a given volume is developed. The results of computational experiments using the developed algorithms are presented. The applicability of the proposed algorithms and approaches is obtained in the framework of the switching node model with recurrent incoming flows, exponential service times, and single-channel transmission lines.

Keywords: switching node; flow management; buffer memory allocation; optimal buffer memory capacity

A HEAD OF THE LINE PRIORITY QUEUE WITH WORKING VACATIONS

• E. S. Kondranin  Department of Mathematical Statistics, Faculty of Computational Mathematics and Cybernetics, M.V. Lomonosov Moscow State University, 1-52 Leninskiye Gory, Moscow 119991, GSP-1, Russian Federation
• V. G. Ushakov  Department of Mathematical Statistics, Faculty of Computational Mathematics and Cybernetics, M.V. Lomonosov Moscow State University, 1-52 Leninskiye Gory, Moscow 119991, GSP-1, Russian Federation, Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The authors analyze the single-server queueing system with two types of customers, head of the line priority, hyperexponential input stream, and working vacations. The authors obtain the Laplace transform (with respect to an arbitrary point in time) of the joint distribution of server state, queue size, and elapsed time in that state. The authors restrict themselves to a system with exhaustive service (the queue must be empty when the server starts a vacation) and multiple vacations. The queueing systems with vacations have been well studied because of their applications in modeling computer networks, communication, and manufacturing systems. For example, in many digital systems, the processor is multiplexed among a number of jobs and, hence, is not available all the time to handle one job type. Besides such an application, theoretical interest in vacation models has been aroused with respect to their relationship with polling models.

Keywords: hyperexponential input stream; working vacations; single server; head of the line priority; queue length

METHODS OF IDENTIFICATION OF HOST CAPTURE IN A DISTRIBUTED INFORMATION SYSTEM WHICH IS PROTECTED ON THE BASIS OF META DATA

• A. A. Grusho   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• N. A. Grusho   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• M. V. Levykin   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• E. E. Timonina   Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The model of a distributed information system in which permissions on network connections are based on meta data is considered. Meta data are simplification of business process models. It is proved that the adversary of information security who captured a host and accurately attacked a system by means of changes of output data of tasks solved on this host cannot be detected at the level of meta data. The problem is connected with the fact that a business process model and, therefore, meta data operate with variables for which changes of specific values are not reflected in their description. Exceptions are output cases on forbidden values, for example, out of limits of definition ranges and a set of values of functions of which information technologies are constructed. Additional variants of information security measures which consider "invisibility" of similar violations of information security at the level of meta data are suggested.

Keywords: information security; information technologies; distributed information system; meta data; ban; threat of host capture

USING A PROBABILISTIC CALCULATION MODEL TO TEST ONE CLASS OF READY-TO-USE SOFTWARE COMPONENTS OF LOCAL AND NETWORK SYSTEMS

• B. M. Basok  MIREA - Russian Technological University, 78 Vernadskogo Ave., Moscow 119454, Russian Federation
• V. N. Zakharov  Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• S. L. Frenkel  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The paper discusses and analyzes the possibility of providing effective testing of ready-to-use software products that solve the task of calculating functions, in the absence of complete information necessary for traditional testing. Efficiency means the possibility of providing an arbitrarily high probability of detecting possible computational errors that were not detected by the output control as the number of inspections increases. As a conceptual model of the proposed approach, the properties of functions with the random self-reducible function are used, that is, functions whose calculation on a particular input set can be reduced to calculation on several randomly selected input sets. The rationality of providing self-testability properties in ready-to-use software is substantiated.

Keywords: software testing; self-testing

SEAMLESS ROUTE UPDATES IN SOFTWARE-DEFINED NETWORKING VIA QUALITY OF SERVICE COMPLIANCE VERIFICATION

• S. L. Frenkel  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• D. Khankin  Computer Science Department, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel

Abstract: In software-defined networking (SDN), the control plane and the data plane are decoupled. This allows high flexibility by providing abstractions for network management applications and being directly programmable. However, reconfiguration and updates of a network are sometimes inevitable due to topology changes, maintenance, or failures. In the scenario, a current route C and a set of possible new routes {Ni}, where one of the new routes is required to replace the current route, are given. There is a chance that a new route Ni is longer than a different new route Nj, but Ni is a more reliable one and it will update faster or perform better after the update in terms of quality of service (QoS) demands. Taking into account the random nature of the network functioning, the present authors supplement the recently proposed algorithm by Delaet et al. for route updates with a technique based on Markov chains (MCs). As such, an enhanced algorithm for complying QoS demands during route updates is proposed in a seamless fashion. First, an extension to the update algorithm of Delaet et al. that describes the transmission of packets through a chosen route and compares the update process for all possible alternative routes is suggested. Second, several methods for choosing a combination of preferred subparts of new routes, resulting in an optimal, in the sense of QoS compliance, new route is provided.

Keywords: software-defined networking; Markov chains; quality of service

OPTIMAL RECURRENT NEURAL NETWORK MODEL IN PARAPHRASE DETECTION

• A.N. Smerdov  Moscow Institute of Physics and Technology, 9 Institutskiy Per., Dolgoprudny, Moscow Region 141700, Russian Federation
• O.Y. Bakhteev  Moscow Institute of Physics and Technology, 9 Institutskiy Per., Dolgoprudny, Moscow Region 141700, Russian Federation
• V.V. Strijov  Moscow Institute of Physics and Technology, 9 Institutskiy Per., Dolgoprudny, Moscow Region 141700, Russian Federation, A. A. Dorodnicyn Computing Center, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 40 Vavilov Str., Moscow 119333, Russian Federation

Abstract: This paper addresses the problem of optimal recurrent neural network selection. It asserts the neural network evidence lower bound as the optimal criterion for selection. It investigates variational inference methods to approximate the posterior distribution of the network parameters. As a particular case, the normal distribution of the parameters with different types of the covariance matrix is investigated. The authors propose a method of pruning parameters with the highest probability density in zero to increase the model marginal likelihood. As an illustrative example, a computational experiment of multiclass classification on the SemEval 2015 dataset was carried out.

Keywords: deep learning; recurrent neural network; neural network pruning; variational approach

AN INFORMATION BASED CRITERION FOR PERFECTLY BALANCED FUNCTIONS

• O.A. Logachev  Institute of Information Security Issues, M. V. Lomonosov Moscow State University, 1 Michurinsky Pr., Moscow 119192, Russian Federation

Abstract: The class of perfectly balanced functions is important for some areas of mathematics, e.g., combinatorics, coding theory, cryptography, symbolic dynamics, and automata theory. It turns out that perfectly balanced functions provide a suitable mathematical tool for description and studying of convolutional codes, cryptographic primitives, surjective endomorphisms of discrete dynamical systems, and information-lossless finite-state automata. Previously, Hedlund and Sumarokov proved criteria of perfect balancedness of functions, which are related to the property of being defect zero and information-lossless. The present author proves a new criterion of the perfect balancedness property in terms of average mutual information. The author also describes a polinomial-time inverting algorithm for perfectly balanced functions.

Keywords: finite alfabet; discrete function; averege mutual information; perfect balancedness; perfectly balanced function; function of defect zero

NEW MIXTURE REPRESENTATIONS OF THE GENERALIZED MITTAG-LEFFLER DISTRIBUTION AND THEIR APPLICATIONS

• V. Yu. Korolev  Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation, Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Hangzhou Dianzi University, Xiasha Higher Education Zone, Hangzhou 310018, China
• A. K. Gorshenin  Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation, Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• A. I. Zeifman  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Vologda State University, 15 Lenin Str., Vologda 160000, Russian Federation, Vologda Research Center of the Russian Academy of Sciences, 56-A Gorky Str., Vologda 160001, Russian Federation

Abstract: The article provides new mixture represenations for the generalized Mittag-Leffler distribution. In particular, it is shown that for values of the "generalizing" parameter not exceeding one, the generalized Mittag- Leffler distribution is a scale mixture of the half-normal distribution laws, classic Mittag-Leffler distributions, or generalized Mittag-Leffler distributions with the larger values of the characteristic index. The explicit expressions for mixing quantities are given for all cases. The obtained representations allow proposing new algorithms for modeling random variables with the generalized Mittag-Leffler distribution and formulating new limit theorems in which such distributions appear as the limit ones.

Keywords: generalized Mittag-Leffler distribution; scale mixture; generalized gamma distribution; half-normal distribution; stable distribution

ON NONUNIFORM ESTIMATES OF ACCURACY OF NORMAL APPROXIMATION FOR DISTRIBUTIONS OF SOME RANDOM SUMS UNDER RELAXED MOMENT CONDITIONS

• V. Yu. Korolev  Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation, Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Hangzhou Dianzi University, Xiasha Higher Education Zone, Hangzhou 310018, China
• A.V. Dorofeeva  Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation

Abstract: Nonuniform estimates are presented for the rate of convergence in the central limit theorem for sums of a random number of independent identically distributed random variables. Two cases are studied in which the summation index (the number of summands in the sum) has the binomial or Poisson distribution. The index is assumed to be independent of the summands. The situation is considered where the information that only the second moments of the summands exist is available. Particular numerical values of the absolute constants are presented explicitly. Also, the sharpening of the absolute constant in the nonuniform estimate of the rate of convergence in the central limit theorem for sums of a nonrandom number of independent identically distributed random variables is announced for the case where the summands possess only second moments.

Keywords: central limit theorem; normal approximation; random sum; binomial distribution; Poisson distribution; Poisson theorem

GAMMA-WEIBULL A PRIORI DISTRIBUTIONS IN BAYESIAN QUEUING MODELS

• E. N. Arutyunov  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• A. A. Kudryavtsev  Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation
• A. I. Titova  Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation

Abstract: This paper considers the Bayesian approach to the queueing theory and reliability theory. Within the framework of the Bayesian approach, it is assumed that the key parameters of classical systems, such as the input flow intensity and the service intensity, are random variables with known a priori distributions. It is reasonable to use the Bayesian approach in the studying of the systems which are of the same type as well as in the studying of one system where changes of its characteristics happen at unpredictable moments of time. The randomization of the system's key parameters leads to the randomization of the system's characteristics, for instance, the traffic intensity In the paper, the analytical results for the traffic intensity probability characteristics in the case of the gamma and Weibull a priori distributions of M/M/1/0 system's parameters are presented. The obtained results are formulated using the gamma-exponential function.

Keywords: Bayesian approach; queuing system; mixed distribution; Weibull distribution; gamma distribution; gamma-exponential function

USING SUPRACORPORA DATABASES FOR QUANTITATIVE ANALYSIS OF MACHINE TRANSLATIONS

• N. V. Buntman  M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation
• A. A. Goncharov  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• I. M. Zatsman  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
• V. A. Nuriev  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The paper discusses an information technology that supports expertise of machine translations. The technology has been developed to meet the following conditions: (i) there are connectives in all translated contexts; (ii) the connectives may be both one-word (khotya 'although,' a 'and') and multiword (da esche 'and beside this,' no zato 'but instead'); and (iii) between words making up a given connective, there may be a space (esli (space) tak 'if (space) then'). With this technology, expertise of machine translations develops through three main stages: (i) linguistic annotation of machine translations in a supracorpora database; (ii) quantitative processing of annotations; and (iii) linguistic analysis of annotations and quantitative data. The paper describes technological aspects of the first two stages. The examples given are only those with multiword connectives. Source sentences chosen for machine translation have been collected from literary texts.

Keywords: supracorpora database; machine translation; classification of errors; technology supporting expertise; linguistic annotation; corpus linguistics; connectives

DEVELOPMENT OF SERVICES OF DIGITAL PLATFORMS TO OVERCOME NONFINANCIAL BARRIERS

• A. K. Gorshenin  Institute of Informatics Problems, Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation, Faculty of Computational Mathematics and Cybernetics, M. V. Lomonosov Moscow State University, GSP-1, Leninskie Gory, Moscow 119991, Russian Federation

Abstract: The article discusses examples of various nonfinancial barriers that impede the effective development of young people in science and education. Overcoming them is not always possible solely by attracting additional resources, since changes in the environment are often required. Conceptual ways to overcome such barriers through creation and development of digital platform services, one of key paradigms in modern information technologies, are proposed. Using the example of the digital platform "Science and Education" created in the Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, the compliance of services with the main directions of the Scientific and Technological Development Strategy of the Russian Federation is demonstrated. The proposed concept is effective for solving a wider class of problems.

Keywords: digital platforms; nonfinancial barriers; strategy of scientific and technological development; digital economy; youth policy