«Systems and Means of Informatics»
Scientific journal
Volume 35, Issue 1, 2025
Content | About Authors
Abstract and Keywords
- I. A. Sokolov
- A. A. Zatsarinny
- V. N. Zakharov
- K. K. Kolin
Abstract: The paper is dedicated to the 90th anniversary of the academician I. A. Mizin - the esteemed scientist in fields of telecommunications, computer science, and information technologies. Here, one focuses on the most significant results he obtained during the 40 years of his scientific work. While working as the chief designer at the Research Institute of Automatic Equipment, I. A. Mizin ensured successful completion of a set of works on the development and implementation of the first domestic geographically distributed data exchange system with packet switching by his team for the purposes of automated control systems of strategic importance. Under his leadership, the research teams of the Institute of Informatics Problems of the Russian Academy of Sciences (RAS) substantiated the choice of information technologies for the creation and development of information networks that comply with the requirements for information security in the interests of the governmental authorities. Igor A. Mizin was a member of the Bureau of the Department of Informatics, Computer Engineering, and Automation of the RAS, was the founder and recognized leader of the domestic scientific school in the theory and practice of information computing and telecommunications systems, and was actively engaged in research and organizational work.
His contribution to the formation and development of the Russian information society is being revealed.
Keywords: chief designer; academician; information and telecommunications networks; information technologies; data exchange system; data switching center; code message switching; data transfer protocols; information security; communication channel; governmental authorities
Abstract: The article is dedicated to the theory of normal suboptimal filters (NSOF) and modified NSOF (MNSOF) for Gaussian continuous and discrete implicit stochastic systems (StS) reducible to explicit. It is supposed that observations do not influence the observable object and are described by explicit stochastic differential and difference equations. A short survey in the field of suboptimal NSOF (MNSOF) synthesis is given. Basic algorithms of NSOF (MNSOF) for the first type synthesis for nonlinear and quasi-linear reducible StS with smooth and nonsmooth implicit nonlinearities are described. The theory of NSOF of the second type for reducible implicit StS is developed on the basis of generalized Kalman-Bucy and Kalman filters. Such NSOF unlike NSOF (MNSOF) of the first type do not permit to estimate accuracy of filtering beforehand applications: quick (or real-time) information processing in technical or organization-technical-economical systems is described by small dimension equations when it is possible to neglect time constants at high derivatives (differences). The results are also applicable to implicit hereditary StS reducible to explicit differential (discrete) StS. Directions for future research are formulated.
Keywords: Gaussian stochastic system (StS); implicit StS; modificated NSOF (MNSOF); normal suboptimal filter (NSOF) of the first and second types
- A. A. Grusho
- N. A. Grusho
- M. I. Zabezhailo
- D. V. Smirnov
- E. E. Timonina
- S. Ya. Shorgin
Abstract: Information about the unavailability or incorrect operation of the service for the user, which is received in the form of a message or complaint to the center for collecting such information, is called an indirect feature of a failure.
Such information is generated by users of services that are not to have the aim in participating in studying the causes and eliminating the failure in the service.
The tasks of analyzing failures in large distributed information infrastructures by indirect features are relevant, since additional objective sources are involved in monitoring the infrastructure - users of services. The article builds a formal graph model of the information infrastructure (hereinafter, the system), which allows one to reflect information about the appearance of failures, taking into account their localization in the system, depending on the information content of indirect features. The conditions were found to detect the root cause of the failure at various levels of the hierarchical description of the system that means the identification of the aggregated object in which the root cause of the failure is located. The examples of propagation of failures and the usage of indirect features to analyze system malfunction during fault propagation are given.
Keywords: failures; analysis of failures by indirect features; cause and effect
- V. N. Gridin
- B. R. Salem
- D. S. Smirnov
- V. I. Solodovnikov
Abstract: Differential privacy (DP) involves adding controlled noise to the input data or computational results which provides robust and provable privacy preservation but can have a significant impact on the statistical datasets properties with potentially influencing effect to their subsequent analysis. This feature forces a trade-off analysis between privacy and utility. The paper presents the research results of the DP Laplace mechanism impact on various data distributions and their statistical properties. The feature changes are studied in the context of three experiments for several key types of distributions while applying the Laplace mechanism with different values of the privacy budget (epsilon). The classical Laplace mechanism is compared with its extensions in the context of their effectiveness and the impact on different levels of feature correlation in the original data is considered. The results highlight the trade-offs between privacy and data utility and also provide recommendations for choosing suitable DP mechanisms for different scenarios.
Keywords: differential privacy; Laplace mechanism; information security; mathematical statistics; correlation; confidentiality; data hiding
- S. A. Stupnikov
- N. A. Skvortsov
- D. O. Briukhov
Abstract: The paper reviews advanced methods for implementing incremental view maintenance in academic and industrial relational database management systems (DBMS). The main directions of materialized views application, the general statement of the incremental maintenance problem, classic methods for view maintenance, basic industrial implementations in the most common DBMS, implementations in modern DBMS, and their distinguishing features are considered. The main directions of academic research, their differences in the problem statement, subsets of the view definition language, and practical or theoretical orientation are considered. Practical-oriented investigations improve industrial DBMS and perform an experimental evaluation of the approach with various scaling options. Theoretical investigations focus on the subsets of the view definition language for which the best update strategies are possible and on the formal proof of the complexity of algorithms. The most andromising methods and ways of implementation in modern DBMS are distinguished.
Keywords: relational database management systems; materialized views; incremental maintenance
- D. O. Dobrovol'skij
- I. M. Zatsman
Abstract: The present paper considers issues of integrating a digital multilingual dictionary (using the example of a German-Russian dictionary) with texts of a parallel corpus within the framework of a lexicographic information system that includes three components: (i) a digital multilingual dictionary; (ii) a corpus as a repository of parallel texts; and (iii) a database of annotated translation correspondences and two knowledge bases. The proposed approach to integration is a synthesis of a number of conceptual procedures, including application of the multilevel structuring principle of dictionary entries, formation of annotated translation correspondences for polysemous words and set phrases along with their translations, and providing links between the digital multilingual dictionary and the repository of parallel texts based on individual meanings of polysemous words and set phrases. Until now, such lexicographic information systems have been developed exclusively for monolingual dictionaries with connecting links by lemmas only. The aim of the paper is to describe the proposed approach to integrating a digital multilingual dictionary with texts of a parallel corpus as a theoretical basis for developing a lexicographic information system.
Keywords: lexicographic information system; parallel texts; digital multilingual dictionary; corpus; database of annotated translation correspondences
- I. A. Kirikov
- S. V. Listopad
- S. B. Rumovskaya
Abstract: The paper is devoted to the issues of application of artificial intelligence systems for solving practical problems characterized by heterogeneity, uncertainty, fuzzyness, and dynamism. The application of decentralized artificial intelligence methods that allow simulating the work of small teams of specialists, which are traditionally used to solve such problems, is substantiated. As one of the possible options for implementing this approach, the concept of hybrid intelligent multiagent systems is proposed, in which intelligent agents with extensive domain models interact with each other, develop rules for jointly solving the problem, exchange knowledge necessary for its solution, adjusting their own ideas about it, distribute its parts among themselves, and synthesize a common solution. This allows the intelligent system to demonstrate high adaptability developing a new relevant method when solving the next problem. The paper proposes a methodology for constructing a typical model and a functional structure of a hybrid intelligent multiagent system, an assessment of existing problems, and possible prospects for the development of research in this area.
Keywords: team of specialists; problem solving; hybrid intelligent multiagent system; construction methodology; conceptual model; functional structure; multiagent system; hybrid intelligent system
- V. N. Zakharov
- Yu. A. Stepchenkov
- Yu. G. Diachenko
- N. V. Morozov
- L. P. Plekhanov
- D. Yu. Stepchenkov
Abstract: The article is devoted to the brief overview of the development trends and achievements of self-timed (ST) circuits. Due to redundant data coding and two-phase discipline, ST circuits have a number of advantages over synchronous counterparts. They operate reliably with any cell delays, detect any constant faults, and have an extremely wide performance range in terms of supply voltage and ambient temperature. Practical development of ST units with varying complexity has proven the effectiveness of ST solutions, especially in highly reliable and fault-tolerant applications. The article presents the comparative test results of synchronous and ST circuit test chips and the performance and fault tolerance estimates for ST circuits of varying complexity. Group indication of multibit ST circuits increases their performance by 40% due to a slight increase (less than 3%) in hardware complexity. The developed design methods of protection against soft errors improve the natural immunity of ST circuits to adverse effects and ensure a level of fault tolerance of ST circuits several times higher than that of the synchronous counterparts.
Keywords: self-timed circuit; C-element; dual-rail; indication; pipeline; performance; fault tolerance
Abstract: The data center processing capabilities have been traditionally implemented by processors of three kinds - central processing units (CPU), graphics processing units (GPU), and network processing units (NPU). Recently, data processing units (DPU), sometimes called also infrastructure processors, have joined this triad. The most frequently advertised purpose of the DPU is to free CPU and GPU from communication burdens and a number of routine tasks for storing, moving, and processing certain data types within a data center that assumes interaction of DPU with CPU and GPU as well as partial or complete replacement of the NPU. The article attempts to clarify the essence of DPU, specifies their place and role in the data center infrastructure, examines the connection of DPU with GPU and NPU, finds out the relation to network functions virtualization, denotes difficulties of using the DPU, and gives a preliminary assessment of the prospects for their further development and application.
Keywords: central processing unit; data processing unit; graphics processing unit; network functions virtualization; network processing unit
|
