ISSN 0236-235X (P)
ISSN 2311-2735 (E)

Journal influence

Higher Attestation Commission (VAK) - К1 quartile
Russian Science Citation Index (RSCI)

Bookmark

Next issue

2
Publication date:
16 June 2024

Articles of journal № 3 at 2012 year.

Order result by:
Public date | Title | Authors

41. Resource distribution in information system related to educational process remote support [№3 за 2012 год]
Authors: Bolodurina I.P., D.I. Parfenov, Reshetnikov V.N.
Visitors: 8559
Development of computer technologies gives rise to the active use of training systems and satellite information technologies in distance education systems. The objective of investigation of effective performance management techniques and optimization of software and hardware usage becomes really relevant. The study is targeted to optimization of resource distribution among vulnerable external (relatively to the system) users in relation to latency and query processing speed. This is connected to availability of internet oriented architecture of the client applications. When the number of queries grows too fast, information systems suffer from lack of consumed resources, there-fore, it is important to manage received queries and minimize failures during the service, which occur depending on allocated resources required for the system operation. Large amount of simultaneous queries leads to disbalance of the operation traffic, negatively affecting routing equipment buffers and results into excessive growth of critical servers’ resources. This study provides construction of multilevel model of the distant education model, it reviews characteristics and provides algorithm for effective use of available resources that can improve service quality in distributed information distant educational systems. Effectiveness of the selection algorithm and service prioritizing of incoming queries is supported by comparative analysis of the output model data, it helps to reduce system time response. In addition, this algorithm can maximize number of processed users’ queries during time unit.

42. Solution of simultaneous polynomial equations on a computer [№3 за 2012 год]
Author: Лёзин И.А.
Visitors: 10065
The article offers the accelerated algorithm of Groebner bases search used in the course of solution of simultaneous polynomial equations. The proposed algorithm considers the issue of word size overflow and incorrect computations in the course of calculating on the computer. The classical Buchberger's algorithm and the updated Faugere's algorithm require a lot of steps; meanwhile they do not contain any additional actions on correction of floating point numbers computations that causes some errors and incorrect computation of simultaneous equations roots and irrational factors. The new algorithm by several times reduces the number of operations of polynom quantity reduction, does not store all possible pairs of polynoms and does not check them for the fact that one can be expressed in terms of the other. The algorithm proposed in the article operates only with that number of polynoms which are now a basis. The basis is expanded if none of the polynoms represented in it can be represented in the form of combination of the other polynoms of this basis. Otherwise, the extra polynom shall be removed from the basis that helps avoid overgrowth of the basis and excess operations with this polynom. In order to avoid the problem of incorrect computations because of the word size of the computer the algorithm envisages post-check. If any of the polynoms is reduced, then the initial polynom should be represented by the combination of the reduced polynom and the rest of the basis polynoms. If the values of coefficients of the equal monoms do not coincide after that, then the coefficients of the reduced polynom require correction by the error size.

43. Role of stochastic testing in microprocessors functional verification [№3 за 2012 год]
Author: Хисамбеев И.Ш.
Visitors: 10222
With the growth of the performance requirements of modern IC, including microprocessors and Systems-on-Chip, their development complicates considerably. It has become a multistage process, and there are many sophisticated tasks to be done on each stage. One of the most labor-consuming tasks is design functional verification. Its goal is to approve a conformity between implementation of a design and it's specification functional requirements. While it does not yet have a general solution, considering modern IC designs complexity, several complementary approaches were developed to address it. One of them is stochastic testing. It was applied for researching of the MIPS64 architecture microprocessors in Science research Institute for system analysis of RAS. The method is based on the test program execution simulation. Test programs are generated automatically from the given template. Instructions, arguments and settings for the test are chosen randomly considering given biases and constraints. This paper is a review, aiming at specifying the role of stochastic testing with its application scope, advantages and disadvantages. In the introduction functional verification is considered in general, as a part of IC design workflow. Then, most well-known verification approaches are reviewed, their underlying ideas analyzed briefly. Particularly, simulation-based methods are considered. Finally, stochastic testing method is described in the given background. Conclusions concerning its advantages and disadvantages are illustrated with some results of its application in SRISA RAS.

44. Samara school of professor S.A. Prokhorov on application analysis of random processes [№3 за 2012 год]
Authors: Ivaschenko A.V., Kulikovskikh I.M.
Visitors: 7470
The article describes the main results obtained at the department of information systems and technologies of Samara State Aerospace University under the guidance of professor S.A. Prokhorov in the field of application analysis of random processes, time series and flows of events. It highlights the issue and describes the main stages of approximative analysis of probabilistic characteristics for a random probabilistic characteristic, for correlation and spectral analysis and for cross correlation and spectral analysis. It briefly specifies the results obtained in the course of research works in the field of development of technology and software for automated systems of random processes application analysis, including: mathematical description, methods and algorithms of simulation of random processes, flows of events and nonuniform time series; methods and algorithms of analysis of distribution laws, characteristic functions, correlation and spectral functions, structural functions; and solution of problems of secondary processing of time series, including identification of random processes in terms of the type of functional characteristic, approximation of distribution laws, characteristic, correlation and structural functions, power spectral density, by means of parametric models, which are the functions of given type as well as the orthogonal functions of exponential type. It gives the description of functionality of the automated systems set enabling to solve different application problems of analysis of random processes and time series. The article also describes the examples of real problems which solution involved the specified methods and algorithms of application analysis of random processes: in physics, acoustics, oceanology, medicine, machine engineering and other fields where the researchers face the necessity to process random processes with various characteristics.

45. Profile sinchronization of MS SharePoint Portal Server 2003 users with external data source [№3 за 2012 год]
Author: Ermakov D.G.
Visitors: 8956
In the process of development of enterprise portals one of most important problems is data synchronization published in such portal with data used by other applications and stored in different formats in DBMS. When the data placed in portal can be edited by a user or other application, they should be synchronized. When one of the data sources assigned as the master source and the other one is assigned as a slave source it is not difficult to arrange synchronization. When bidirectional structure is needed, it shall be necessary to provide «direct» and «reverse» synchronization. The article discusses synchronization solution of MS SharePoint Portal Server 2003 (SPS) users with a data source – inherited personnel department subsystem and «reverse» synchronization of this external source with profile data of SPS users. The article presents two ways of data collection from user profiles: with SQL query and use of object SPS model. Synchronization is made with temporary XML file. Such solution changes source of synchronized data without significant change in existing software. In addition, it provides opportunity to the third party to transfer or receive the data from it. The article contains some scripts written in PowerShell scenario language included in standard MS Windows package with implementation of «direct» and «reverse» synchronization. Beside synchronization, this approach can be used for the data transfer in case of changing MS SharePoint platform to a different one.

46. Planning system for production with concurrent product assembly [№3 за 2012 год]
Authors: Сидоренко А.М., Khobotov E.N., Мелкишев В.Н.
Visitors: 10257
The article offers ideas and algorithm of planning and scheduling of works at the enterprise level. There is considered the computational algorithm scheme to make schedules for components processing and units assembly in mechanical workshops of an enterprise, as well as the structure of the planning and scheduling system for enterprises, created using the proposed algorithm. To verify functionality of the proposed algorithms and software prototype of the system there was held a series of numerical experiments showed their sufficiently high effectiveness. Creation of planning and scheduling systems for enterprises on the basis of such algorithms can significantly improve the efficiency of production.

47. Contemporary issues of practical use of multiagent systems [№3 за 2012 год]
Authors: Иноземцев С.В., Дмитриев А.Ю.
Visitors: 10730
This article describes the current issues of building and practical use of multiagent systems and architectures which programmer-designers face in the course of solution of practical problems in various subject fields. Various viewpoints on multiagent platform and peculiarities of its usage in the course of building of intelligent software were outlined on the basis of contemporary understanding of mutliagent architecture and its role in practical applications. The hybrid solution based on combination of intelligent and classical algorithms of optimization to build automated resource management system in transport logistics is described. Solution MAXOPTRA is given as an example; it ensures complex management of orders and shipping in the forwarding company working in the field of cargo-and –passenger transportation and the results of complex automation of transport planning processes at regional branches of ROSINKAS. The article reviews application of multiagent technologies in modeling problems through the example of the hotel management problem; in addition, parallels were drawn between approaches on organization of management system in transport logistics and organization of hotel services management, which allowed the possibilities of unifying of multiagent approaches to be highlighted. As a result, the common basis for a set of business issues was emphasized and architecturally formalized in the form of multiagent platform: basic algorithms include orders consolidation on resource, «constraint stressing», planning of bound schedules of several resources. Keywords: software platforms, multi-agent technology, resources management.

48. Creation of software system on the basis of spatial object interaction diagram [№3 за 2012 год]
Authors: Prokhorov S.A., Kulikovskikh I.M.
Visitors: 10299
The article looks into the possibility of software system creation for data analysis within Data Mining systems with a staged platform of data analysis. In this regard a new solution is provided for software products creation focused on incremental processing of the initial information in the form of a spatial object interaction diagram, which will allow time and resource expenditures required to obtain the final software product to be reduced. The developed spatial diagram is built on the basis of the following fundamental notions: functional transformation space, functional enhancement space, characteristic enhancement space, key of transition from one space to the other at multiple levels. On the basis of the developed spatial diagram the structural diagram of software system for correlation and spectral data mining was created, while this kind of analysis, according to the research at the market of the analogous products, is unique in terms of the class of current problems. Mathematical models taken as a basis for this software product are based on Fourier orthogonal decomposition of functional characteristics, and for implementation of a hidden level of data analysis the analytical processing technique was created, which consists in functioning management and implementation of created search mechanisms for hidden regularities. Using spatial object integration diagram reduces the labour intensity of software development for automated systems of scientific research, control and data processing, and can be recommended to the designers of automated systems – programmers, as well as to the analysts involved in designing of complex software systems.

49. Improvement of debugging and testing increase performance techniques in multiprocessor systems [№3 за 2012 год]
Author: Lavrinov G.A.
Visitors: 11091
Very important element of multiprocessor computing system is communication network used by process to provide communication with other processors or memory. Together with VME, PCI Express, HyperTransport and other buses of inter-processor communication, RapidIO interface is developed as well. Microprocessor systems based on RapidIO suffer significant difficulty and present a serious problem at debugging and beginning testing stage of trial and developmental models. Full scale testing is made with operating systems (in this case – Linux and OS PB Baget 2.0 and 3.0), however, successful start of operating systems, communication network and processor units are supposed to provide proper operating performance. Test designer has only hardware and ROM program that gets control automatically after power is on or RESET signal is received. This article provides two testing and debugging techniques of multiprocessor systems realized with RapidIO interface that can work with minimum amount of hardware. The article also provides comparison of these techniques viewing effectiveness in use and testing stage made on its basis. Using UML sequence chart, the article presents protocol that implements integrated RapidIO console and I/O RapidIO communication protocol for gathering information about testing results. It is shown how to use specific RapidIO packages. These testing techniques of multiprocessor systems became a basis for performing system test based on 1890BM6YA chips .

50. Verification of a microprocessor and its RTL-model by means of ОС Linux user applications [№3 за 2012 год]
Author: Chibisov P.A.
Visitors: 11220
This article covers the methods of verification and testing of modern microprocessors. The special attention is given to a method of testing the RTL-models, FPGA-prototypes and test crystals of microprocessors by real user applications for the Linux operating system. The interrelation of these objects and degree of a discussed technique applicability to each of them in a context of the general verification plan are considered either. The article lists the merits and demerits of the method. As simulation speed of programs on RTL-model of a microprocessor is extremely slow, it is offered to use the cut-restore mechanism of a model state for splitting all instruction sequence of an operating system booting into a set of subsequences which are carried out in parallel on different computers. Existence of a large quantity of freely distributed programs with an open source code with the built-in automated self-test mechanisms makes it possible to mark out an applications launch for Linux OS in a separate approach for testing of universal microprocessors. The described method doesn't exclude, but serves as a supplement to a modern set of methods and means of testing and verification of microprocessors and their models. Many authoritative developers and manufacturers of microprocessors recognize the usefulness of the earlier booting of any operating system on the RTL-model under development. Success in this operation often makes the developers more confident that their work is done correctly, than tens of thousands of executed tests. The article provides the example of the representative test set which makes it possible to use the ready-made user software packages, as well as the examples of the test program sources. Besides, it considers the general algorithm of actions to find a bug in the microprocessor and gives examples of bugs revealed in the microprocessor with MIPS64 architecture.

← Preview | 1 | 2 | 3 | 4 | 5 | 6 | Next →