ISSN 1727–6209

TERNOPIL NATIONAL ECONOMIC UNIVERSITY RESEARCH INSTITUTE OF INTELLIGENT COMPUTER SYSTEMS

ТЕРНОПІЛЬСЬКИЙ НАЦІОНАЛЬНИЙ ЕКОНОМІЧНИЙ УНІВЕРСИТЕТ

НАУКОВО-ДОСЛІДНИЙ ІНСТИТУТ ІНТЕЛЕКТУАЛЬНИХ КОМП’ЮТЕРНИХ СИСТЕМ

International Journal of

Computing

Published since 2002

December 2007, Volume 6, Issue 3

Міжнародний науково-технічний журнал

Комп’ютинг

Видається з 2002 року

Грудень 2007, Том 6, Випуск 3

Постановою президії ВАК України № 1-05/10 від 10 грудня 2003 року науково-технічний журнал “Комп'ютинг” віднесено до переліку наукових фахових видань України, в яких можуть публікуватися результати дисертаційних робіт на здобуття

наукових ступенів доктора і кандидата технічних наук

Тернопіль “Економічна думка”

2007

“Computing” December 2007, Vol. 6, Issue 3

International Journal of Research Institute of Intelligent Computer Systems, Ternopil National Economic University

Registered by State Committee on Information Policy, Television and Broadcasting of Ukraine. Registration certificate – КВ №11095 of 28.02.2006.

Editorial Board

EDITOR-IN-CHIEF Anatoly Sachenko Ternopil National Economic University, Ukraine EDITOR-IN-CHIEF DEPUTY Oleg Berezsky Ternopil National Economic University, Ukraine ASSOCIATE EDITORS Mykola Cherkaskyy National University “Lviv Polytechnics”, Ukraine Nikos P. Chrisochoides William & Mary College, USA Pasquale Daponte University of Sannio, Italy Mykola Dyvak Ternopil National Economic University, Ukraine Richard J. Duro University of La Coruna, Spain Vladimir Golovko Brest State Polytechnical University, Belarus Sergei Gorlatch Technical University of Berlin, Germany Lucio Grandinetti University of Calabria, Italy Dan Grigoras University College Cork, Ireland Halit Eren Curtin University of Technology, Australia Vladimir Haasz Czech Technical University, Prague, Czech Republic Robert Hiromoto University of Idaho, USA Sitharama Iyengar Louisiana State University, USA Mykola Karpinsky Ternopil National Economic University, Ukraine

Yury Kolokolov State Technical University, Orel, Russia Janusz S. Kowalik The Boeing Company, USA Volodymyr Kozhemyako Vinnitsa State Technical University, Ukraine Gennady Krivoula Kharkiv State Technical University of Radioelectronics, Ukraine Natalia Kussul Space Research Institute, Ukraine Theodore Laopoulos Thessaloniki Aristotle Univerity, Greece Viktor Lokazyuk Technological University Podillya, Ukraine Fernando López Peña Universidade da Coruña, Spain Kurosh Madani Paris XII University, France Prabhat Mahanti University of New Brunswick, Canada George Markowsky University of Maine, USA Anatoly Melnyk National University “Lviv Polytechnics”, Ukraine Mykola Nedashkovskyy Ternopil National Economic University, Ukraine Yaroslav Nykolaiychuk Ternopil National Economic University, Ukraine Oleksandr Palahin V.M.Glushkov Institute of Cybernetics, Ukraine. Lyubomyr Petryshyn Ternopil National Economic University, Ukraine Vincenzo Piuri Politecnico di Milano, Italy John Pollard University College London, UK

Mykola Pryymak Ternopil State Ivan Puluy Technical University, Ukraine Peter Reusch University of Applied Sciences, Dortmund, Germany Sergey Rippa State Tax Service Academy of Ukraine, Irpin, Ukraine Volodymyr Romanov V.M.Glushkov Institute of Cybernetics, Ukraine Bohdan Rusyn Physical and Mechanical Institute of Ukrainian National Academy of Sciences, Ukraine Rauf Sadykhov Byelorussian State University of Informatics and Radioelectronics, Belarus Valeriy Shyrochyn National Technical University of Ukraine “Kyiv Polytechnics Institute”, Ukraine Tarek M. Sobh University of Bridgeport, USA Petro Stakhiv National University “Lviv Polytechnics”, Ukraine Volodymyr Tarasenko National Technical University of Ukraine “Kyiv Polytechnics Institute”, Ukraine Grygoriy Tsegelyk Lviv Ivan Franko National University, Ukraine Allen B. Tucker Bowdoin College, USA Alain Vande Wouwer Faculte Polytechnique de Mons, Belgium Bogdan M. Wilamowski University of Idaho, USA Wieslaw Winiecki Warsaw University of Technology, Poland Hideki Yamamoto Okayama University, Japan

Address of theResearch Institute of Intelligent Computer Systems Ternopil National Economic University 3, Peremoga Square Ternopil, 46004, Ukraine

Editorial Board Phone: +380 (352) 43-60-38 Fax: +380 (352) 43-63-54 computing@tanet.edu.te.ua www.computingonline.net

© International Journal of Computing, 2007

“Комп’ютинг” Грудень 2007, том 6, випуск 3

Міжнародний науково-технічний журнал Науково-дослідного інституту інтелектуальних комп’ютерних систем Тернопільського національного економічного університету

Зареєстрований Державний комітетом інформаційної політики, телебачення та радіомовлення України. Свідоцтво про реєстрацію – КВ №11095 of 28.02.2006.

Редакційна колегія

ГОЛОВНИЙ РЕДАКТОР Анатолій Саченко Тернопільський національний економічний університет, Україна ЗАСТУПНИК РЕДАКТОРА Олег Березький Тернопільський національний економічний університет, Україна ЧЛЕНИ РЕДКОЛЕГІЇ Микола Дивак Тернопільський національний економічний університет, Україна Владимир Головко Брестский государственный политехнический университет, Беларусь Микола Карпінський Тернопільський національний економічний університет, Україна Володимир Кожем’яко Вінницький державний технічний університет, Україна Юрий Колоколов Государственный технический университет, Орел, Россия Геннадій Кривуля Харківський державний технічний університет радіоелектроніки, Україна Наталія Куссуль Інститут космічних досліджень, Україна Віктор Локазюк Технологічний університет Поділля, Хмельницький, Україна Анатолій Мельник Національний університет “Львівська політехніка”, Україна Микола Недашковський Тернопільський національний економічний університет, Україна Ярослав Николайчук Тернопільський національний економічний університет, Україна Олександр Палагін Інститут кібернетики ім.

В.М.Глушкова НАНУ, Київ, Україна Любомир Петришин Тернопільський національний економічний університет, Україна Микола Приймак Тернопільський державний технічний університет ім. Івана Пулюя, Україна Сергій Ріппа Академія Державної податкової служби України, Ірпінь, Україна Володимир Романов Інститут кібернетики ім. В.М.Глушкова НАНУ, Київ, Україна Богдан Русин Фізико-механічний інститут НАНУ, Львів, Україна Рауф Садыхов Белорусский государственный университет информатики и радиоэлектроники, Минск, Беларусь Петро Стахів Національний університет “Львівська політехніка”, Україна Володимир Тарасенко НТУУ “Київський політехнічний інститут”, Україна Григорій Цегелик Львівський Національний університет ім. Івана Франка, Україна Микола Черкаський Національний університет “Львівська політехніка”, Україна Валерій Широчин НТУУ “Київський політехнічний інститут”, Україна Nikos P. Chrisochoides William & Mary College, USA Pasquale Daponte University of Sannio, Italy Richard J. Duro University of La Coruna, Spain Sergei Gorlatch Technical University of Berlin, Germany Lucio Grandinetti University of Calabria, Italy

Dan Grigoras University College Cork, Ireland Halit Eren Curtin University of Technology, Australia Vladimir Haasz Czech Technical University, Prague, Czech Republic Robert Hiromoto University of Idaho, USA Sitharama Iyengar Louisiana State University, USA Janusz S. Kowalik The Boeing Company, USA Theodore Laopoulos Thessaloniki Aristotle Univerity, Greece Fernando López Peña Universidade da Coruña, Spain Kurosh Madani Paris XII University, France Prabhat Mahanti University of New Brunswick, Canada George Markowsky University of Maine, USA Vincenzo Piuri Politecnico di Milano, Italy John Pollard University College London, UK Peter Reusch University of Applied Sciences, Dortmund, Germany Tarek M. Sobh University of Bridgeport, USA Allen B. Tucker Bowdoin College, USA Alain Vande Wouwer Faculte Polytechnique de Mons, Belgium Bogdan M. Wilamowski University of Idaho, USA Wieslaw Winiecki Warsaw University of Technology, Poland Hideki Yamamoto Okayama University, Japan

Адреса редакції :Науково-дослідний інститут інтелектуальних комп’ютерних систем Тернопільський національний економічний університет площа Перемоги, 3 Тернопіль, 46004, Україна

Тел.: +380 (352) 43-60-38 Факс: +380 (352) 43-63-54 computing@tanet.edu.te.ua www.computingonline.net

© Міжнародний науково-технічний журнал “Комп’ютинг”, 2007

Contents / Computing, 2007, Vol. 6, Issue 3, 4-5. Зміст / Комп’ютинг, 2007, Том 6, Випуск 3, 4-5

4

CONTENTS

A. Amrouche, A. Taleb-Ahmed, J. M. Rouvaen, M. C. E. Yagoub A Robust Speech Recognition System Using a General Regression Neural Network.................................... 6

A. Gladun, J. Rogushina Formalization of Search Context on Base of Ontologies and Multilinguistic Thesauruses ............................ 16

F. H. Elfouly, M. I. Mahmmoud, M. I. M. Dessouky, S. Deyab Comparison Between HAAR And Daubechies Wavelet Transformions on FPGA Technology.................... 23

П. Илиев, П. Цветков, Г. Петров Многомерный aнализ динамических сцен. Прослеживание движущихся объектов в последовательностях изображений ............................................................................................................. 30

A. Cheptsov Simulation Service Providing in a Distributed Simulation Environment........................................................ 38

G. Popov Failures Detection Methodology in non Recovery Computer Systems Based on Diversity Modeling........... 46

В. П. Кожем’яко, А. А. Яровий, Р. М. Новицький Нанотехнологічні принципи реалізації оптоелектронного модуля для запису, збереження та відображення інформації .............................................................................................................................. 52

A. Amari, N. El Bari and B. Bouchikhi Electronic Nose for Anchovy Freshness Monitoring Based on Sensor Array and Pattern Recognition Methods: Principal Components Analysis, Linear Discriminant Analysis and support vector machine ........ 61

V. Nikulshin, V. von Zedtwitz Optimal Disign of Energy Supply Nets on Graphs ......................................................................................... 68

S. A. Khan, T. Islam, Gulshan Artificial Neural Network Based Online Sensor Calibration and Compensation ........................................... 74

M. O. Vakulenko Acoustic Invariant Approach to Speech Sound Analysis for Brand New Speech Recognition Systems (Ukrainian and English) .................................................................................................................................. 79

K. P. Vidya Distributed Trust in ePayments System........................................................................................................... 87

S. V. Dzubin, A. V. Matsiuk, S. V. Martsenko, M. V. Pryimak The Choice and Substantiation of the Mathematical Model of Electroretinogram in the Form of Linear Stochastic Process ........................................................................................................................................... 95

A. Rathinavelu, H. Thiagarajan Computer Aided Articulatory Tutor: a Scientific Study..................................................................................100

З. Домбровський Оцінка кількості ідентифікованих об’єктів системи керування за тональним сигналом.......................106

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

38

SIMULATION SERVICE PROVIDING IN A DISTRIBUTED SIMULATION ENVIRONMENT

Alexey Cheptsov

Donetsk National Technical University, 83000, Donetsk, vul. Artema 58,

e-mail: lex@cs.dgtu.donetsk.ua, www.donntu.edu.ua

Abstract: Simulation of complex dynamic systems is an interdisciplinary problem. To solve this problem the author suggests a distributed simulation environment as a new form of simulation means system organization that provides web-based distributed simulation of engineering and industrial tasks as simulation services using Discrete Event Simulation Framework on the basis of Grid infrastructure. The Unified Process of Simulation Services Development based on Unified Modeling Language is proposed. Keywords: Distributed simulation environment, system organization, simulation service providing.

1. INTRODUCTION Internet as a central communication environment

and progress of information technologies for the last 10 years have had a strong influence on the simulation technology and especially on web-based simulation that is an important research direction at the area of Discrete Event Simulation (DES) [1,2,3]. At the same time the integration of Application Service Providing into Business-to-Business Communications has given the software developers a functional possibility of Service Providing also for tasks of industrial simulation [4,5].

The classical Application Service Providing is founded on interactions between user and computer (computer application). Development of XML and related technologies has allowed data exchanges between separate software components by means of standard communication protocols and interfaces and acts as a basic point in wide-spreading simulation resources into the internet infrastructure based on Web Services technology, Grid computing and Simulation Service Providing. Such standardization allows functional transforming of the Internet as an Information Grid to a Service Grid, in that information and application resources are being requested both by users and by application (software components) transparently [6].

Organization of simulation as an Application Service Providing (Simulation Service Providing, SSP) requires new approaches of simulation means system organization. A new generation of simulation tools is being developed no more as simple software tools, but as distributed simulation environments

those are an infrastructure for integration of software systems and concentrate the most modern engineering and information-technical software solutions at the area of simulation technology. This requires special developing frameworks based on Service Oriented Architecture (SOA). The clients in such frameworks are connected to the server environment through the standard internet protocols (Fig. 1).

Outputin chosen Form(text, diagram)

SSP-Client

Model traversingin XML

Simulation Optional module

SOA

P/H

TTP

Modelelaboration/configuration

(GUI/Text)

Response generation

SSP-Server

Model recognizing

Fig. 1 – Schematic construction of

SSP-Proceeding.

computing@tanet.edu.te.ua www.computingonline.net

ISSN 1727-6209 International Scientific

Journal of Computing

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

39

Grid-Computing [7] – a term that means a Computer-, Network- and Software-Infrastructure of resources distribution with the aim of virtual organization tasks solving – allows to develop Simulation Services in distributed simulation environments on a basis of Open Grid Service Architecture (OGSA). OGSA technologies make possible to connect within a single communication and developing framework many SSP-Clients and SSP-Servers as shown in Fig. 1 with further distribution of Grid-Services between Server-Resources, what takes use of them to a perspective direction of simulation technology’s development and is an important progress in experience accumulation for solving engineering problems.

Herewith the following main requirements are to be taken into account: preparing user-friendly interactive dataset in suitable for engineers graphical and numerical form; integration with existing systems of industrial automation; models correspondence to specifications of problem domain’s industrial processes with an understandable and engineers-oriented representation of simulation study; models providing with special training simulators for operators.

The technological potential of Grid infrastructure is imperative for scientific progress. It enables simulation researches and visualization by solving important industrial tasks those would otherwise remain undiscovered and unexplained – e.g. in climate research or material science. Implementation of new approaches of SSP helps researches carry out detailed simulations and are extremely valuable for industry as well because of enabling noticeable reduction in development’s and practical use’s costs.

2. OBJECTS AND TASKS OF THE SIMULATION SERVICE PROVIDING.

COMPLEX DYNAMIC SYSTEMS. The goal of Grid-based SSP is to build a large

scale resource sharing infrastructure that enables easy and transparent access of engineers to simulation resources, calculating various simulation sub-tasks on the most suitable computing architecture and subsequently combining the results. It aims also at speeding up scientific simulations for such application areas as for example, weather prediction, astrophysics, bioinformatics, aerodynamics, earth quake research, ground water pollution, multi particle physics and other. All this application areas can be defined by term of Dynamic Systems (DS) – technological objects in those the controlled processes of states modification occur. The general similarities of investigated DS of different application areas are a complex, nonlinear, multimodal objective function, the number of state

points those could be used within a single study, large set of parameters and constraints of their simulation models. Such objective function’s behavior requires solving global optimization tasks. Let us consider some examples of typical DS-simulation tasks for different application areas.

Aerodynamics. Until now it is necessary to carry out complicated, coupled simulations of air- and gas- flows distribution in complex aerodynamic objects those are frequently used for providing safety of working conditions. A typical example of such objects is a mine ventilation network. The main fragment of a mine ventilation system is an extraction area’s ventilation system (Fig. 2).

X

Z

Y

Q1

Q3

q q

Q2

Extra

ction

Drif

t

Venti

lation

Drif

t

Lava

Extraction Area

Fig. 2 – The structure of an extraction area’s ventilation system.

The simulation model of an extraction area’s

ventilation system is formed by equations of an Extraction Drift

⎪⎪⎩

⎪⎪⎨

⎧

+∂∂

=∂∂

−

++∂∂

+∂∂

=∂∂

−

,

;)(2

1

21

1

21

211

211

1

1

112

1

1

gFсa

жQ

Fсa

tp

QtrQrt

QFс

жQQ

Fс

жp

(1)

an approximated by parallel flows Extraction Area

⎪⎪

⎩

⎪⎪

⎨

⎧

∂∂

=∂

∂−

+∂∂

+=∂

∂−

,

;

2

2

жg

mFсa

tp

gglFс

tg

mFсg

kFсх

жp

g

g

ggg

g

(2)

a Lava

⎪⎪⎩

⎪⎪⎨

⎧

∂∂

=∂∂

−

∂∂

+=∂∂

−

,

;

2

2

22

2

2

222

2

жQ

Fсa

tp

tQ

FсQr

жp

(3)

a Ventilation Drift

⎪⎪⎩

⎪⎪⎨

⎧

−∂∂

=∂∂

−

′++∂∂

+∂∂

=∂∂

−

,

;2

3

23

3

23

233

233

3

3

332

3

3

gFсa

жQ

Fсa

tp

QrQrt

QFс

жQQ

Fс

жp

(4)

and boundary conditions [8]. Here p1,2,3,g – pressures; Q1,2,3,,g – airflows; ζ –

space coordinates, ρ – air density; а – speed of

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

40

sound in air; F1..g – cross section’s area; r1..g – equivalent specific resistance; r(t), r ′ – variable aerodynamic resistances in drifts; l1..g – length of an element; υ – cinematic viscosity index, k – penetration rate, m – rate voids.

Solving of referred to mine ventilation systems simulation tasks is the objective of scientific researches and developments of the Donetsk National Technical University (Ukraine) in cooperation with High Performance Computing Technology Center of Stuttgart (University of Stuttgart, Germany).

Molecular Dynamics. Classical Molecular Dynamics is used to simulate the dynamics of large systems, to gain a deeper understanding of the interactions between the atoms. It has an application in various scientific and engineering areas such as: chemical and process engineering, turbulent flows and process engineering, bio and enviromental processes and other. One example is the science test bed projects funded by Department of Earth Sciences of University of Cambridge [9] with the goal of primarily using grid-computing methods to facilitate simulations of environmental processes at a molecular level with increased levels of realism, achieved through being able to use larger samples, more complex systems, or through increased accuracy to better handle changes in chemical bonding.

3. THE DISTRIBUTED SIMULATION ENVIRONMENT (DSE) FOR COMPLEX

DYNAMIC SYSTEMS With the aim of virtual simulation models of

different research areas unified development, their de-virtualization and decomposition into subsystems, distributed simulation service providing in scientific, industrial, business and educational researches [10] a Distributed Simulation Environment (DSE) is proposed (Fig. 3), which is a form of hardware, system software and simulation software resources system organization that supports all phases of models development, implementation and industrial use in order to requirements defined above. Hardware resources of DSE contain besides usual user work places also high-performance computer systems of SIMD- and MIMD-Architectures and clusters of nonparallel PCs. The Hardware of periphery systems is used for training simulators and their communication systems production, visualization, as well as for documentation of DSLB- and DSDP-simulation results. System Software of DSE consists of SIMD- and MIMD-based parallel and distributed operation systems, software platforms of parallel and distributed programming, system means of periphery organization and functions providing.

DistributedSimulationEnvironmentfor complexdynamicsystems

Hardware Resources Simulation Software SystemSoftware

High - performance computer resources ( SIMD , MIMD ,

ParCLUSTER )

Modeling - oriented periphery systems

Work place 1

Work place N

Parallel operationsystem;Platformsof parallel

programming

Distributedoperationsystemsand programmingplatforms

System organizationofperiphery

Parallel simulation Softwarefor lumped - parameter

dynamic systems ( DSLB - Simsoftware )

Parallel simulation Softwarefor distributed parameter dynamic systems ( DSDP -

Simsoftware )

Visualisation Software (VisualSimsoftware )

Userinterface Subsystem

System part Model - oriented part

Subsystem of user’ s access to resources

DistributedSimulationEnvironmentfor complexdynamicsystems

Hardware Resources Simulation Software SystemSoftware

High - performance computer resources ( SIMD , MIMD ,

ParCLUSTER )

Modeling - oriented periphery systems

Work place 1

Work place N

Parallel operationsystem;Platformsof parallel

programming

Distributedoperationsystemsand programmingplatforms

System organizationofperiphery

Parallel simulation Softwarefor lumped - parameter

dynamic systems ( DSLB - Simsoftware )

Parallel simulation Softwarefor distributed parameter dynamic systems ( DSDP -

Simsoftware )

Visualisation Software (VisualSimsoftware )

Userinterface Subsystem

System part Model - oriented part

Subsystem of user’ s access to resources

Fig. 3 – Structure organization of Distributed Simulation Environment for complex lumped-parameter and distributed parameter dynamic systems.

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

41

Modeling and Simulation Software contains three main components – for DSLB- and DSDP-simulation and for providing visualization, those implement functionality of basic integrated in DSE simulation components which realize the unified algorithm of simulators development [11] – topological analyzer (intended for preparing discreted primary topological information that is necessary for developing simulation models), equations generator (builds a set of equations systems on the basis of prepared by topological analyzer data and transforms them into virtual simulation models in conventional form) and equations solver (aims at cyclic realization of an algorithm of numerical solving generated equation systems in matrix-vector form, is developed separately for different numerical methods) [12].

With aim of systematization of the models development a DSE-decomposition into following subsystems is proposed [13]:

- user interface subsystem (consists of system- and modeling-oriented parts);

- subsystem of solving parallel equations; - information transferring subsystem; - processors load-balancing subsystem; - data organization subsystem; - subsystem of user access. Developing of defined subsystems oriented on

features of dynamic systems of different application areas like shown above results in implementation of technologically-oriented DSE those implement the accumulated experience of performing simulation tasks in such application areas on the principally new methodical and technological level.

4. THE SYSTEM ORGANIZATION OF TECHNOLOGICALLY-ORIENTED

DISTRIBUTED SIMULATION ENVIRONMENT FOR SOLVING

INDUSTRIAL TASKS Organizationally a technologically-oriented DSE

represents a software-technical infrastructure that provides users (engineers and operators of DSE-application area, as well as researchers and simulators developers) with services those are foreseen within simulation service providing.

Within this infrastructure the following classes of DSE-resources can be defined (Fig. 4):

1. Workplaces of SSP-users (interconnected workplaces of specialists in predefined application area, computing centers, industrial automation systems).

2. Computing resources of problem domain’s simulation tasks analysts (scientific and researching organizations – specialists in model-driven DS-supporting).

Fig. 4 – The organizational structure of a technologically-oriented distributed simulation

environment.

3. Resource of HPC-owners – industrial and scientific organizations those are taking part in simulation researches supporting by their high-performance computer environments.

4. Resources of decision-making’s supporting systems of experts and controlling organizations those specialize in predefined application area.

This resource sharing infrastructure might aim at the achieving the highest possible performance in solving widespread industrial simulation tasks.

The system organization of technologically-oriented DSE includes 3 basic layers of hardware and software resources (Fig. 5):

- Users layer - Server layer - Supercomputer layer

This SimGrid-system organization foresees easy and transparent accessing heterogenous user resources (e.g. PCs, laptops, mobile devices etc.) to DSE-simulation services providing uniform high-performance computational facilities. Models management, simulation jobs remote submission and monitoring, Grid-resource reservation, secure data transport and another user tasks performe through the web-interface which is specially developed for different application areas and hierarchical types of users within each specified application area.

The main administrative domain of DSE is the SimGrid-Server, the main functional components of that (models server, internet server, database server) service the high-level functionality of SimGrid-System – users authorization and authetification, simulators management, provide user access to simulation services. The simulators are developed as Grid-Services according to OGSA-standard [14] with further distribution of simulators set within high performance computing environment of supercomputer layer, the resources of which belong to different administrative domains, e.g., different research institutions in different countries or different companies.

IIS-Server HUBDataBase-Server

A dministratorSpezialist des BRDs

Hauptserver

Internet

Arbeitsplatz1

Arbeitsplatz2

A rbeitsplatz3

Klientennetz

HLRS NEC Paragon CLiCCluster IPVS

High-performance resources UniversitiesUsers

Main Server

Experts Hardware

Uni-ErlangenUni-MagdeburgUni-Stuttgart

Administrators Hardware

NetworkWork place

n

Work place 1

Server of Industrial

Automation

Administration System

Internet-Server

Database-Server

Modeling-Server

1 2

3

4

Expert's Hardware

Internet

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

42

5. SIMULATION SERVICES DEVELOPMENT IN DSE-DESCRETE EVENT SIMULATION FRAMEWORK Developing and providing simulators as

application specific services requires a discrete event simulation framework that meets the defined for DSE requirements and provides necessary level of quality-of-service as well. This framework is tightly coupled with developments of DSE-components, what causes their orientation not only on simulation service but also on development tasks. With this aim special problem-oriented Simulation and Servicing Centers (SSC) could be proposed [15] that integrate DSE for specified application area, as well as software technologies and platforms of their development. Furthermore, development of SSC aims at practical providing defined subsystems of technologically-oriented DSE for scientific and industrial on-line simulation in suitable for experts of specified application area user-friendly form.

In this context an especial role obtains using progressive informational technologies that aim at organization and supporting of software components development that realize simulators functionality according to principals of DSE system organization.

5.1. Using Unified Modeling Language (UML) for the Development of Software Components of DSE-Simulators

The complexity of DSE-components and function algorithms of their implementation within defined SSC’s functionality, the distributed nature of technologically-oriented DSE’s developers leads to a necessity of using a special technology of component-based software systems development. In connection to this an especial interest awakes choosing a special notation of SSC-projects definition, modeling components development’s management, supporting and documentation as well. As a possible unified solution for all tasks of simulators development and further supporting, system integration within another SSC’s service and modeling components could be proposed UML-based technologies. UML [16] is a wide-spread modeling language for software and systems development which became last years practically a standard of object-oriented simulation. The UML can be effectively used for solving following DSE’s development tasks: logical notion of researched objects, component-based simulators development, using simulators for solving industrial tasks, DSE’s system organization, simulation study’s procedure.

Fig. 5 – The system organization of Grid-based distributed simulation environment.

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

43

Complex Dynamic System (DS)

Гірничасправа

Activity Diagram

SimulatorsDiagram

Specified application

area Diagram

Specified application

area Diagram

Data Diagram

Sequence Diagram

Classes Diagram

Захистдовкілля

UML-Model

Deployment Diagram

Timing DiagramU

ML-

base

d D

escr

iptio

nFo

rmal

Des

crip

tion

Requirements managing

Simulators development phases

System analysis

Projecting

Realization4

1

2

3

11

2

2 3

4

4

4

Coa

l ind

ustry

App

licat

ion

area

s of S

imul

atio

n Se

rvic

e Pr

ovid

ing

Hyd

raul

icEl

ectri

cal e

ngin

eerin

g…

? ???? ??????? ????Topological Description

Systems of usual differential equations

Elliptic equations

Parabolic equations

Hyperbolic equations

Special partial differential equations

Technological schemas

Graphs

Structural schemas of industrial automation

? ???? ??????? ????Mathematical Description

Semantic networksSystems of partial

differential equations

Systems of algebraic equations

Fig. 6 – The unified process of simulation services development for complex dynamic systems.

5.2. The Unified Process of Simulation Services Development

With the aim of simulation services development a based on integration of Rational Unified Process [17] with Waterfall Model, tightly coupled with UML technologically-oriented, Use-Case-Driven, iterative and architecture centric unified process is proposed (Fig. 6) that includes the following development phases: requirement managing, system analysis, projecting, realization. Each development phase is performed by means of defined set of UML-diagrams development. Herewith the actual task was adapting defined phases to peculiarities of SSP in the most common application areas of simulation technology (e.g. coal industry, hydraulic, electrical engineering) and to the features of proper to them dynamic systems formal description which contains two basic parts – mathematical description and topological description. The adoption was performed through the developing a new diagrams set that is not foreseen within UML 2.0 standard [16].

Based on experiences for real industrial projects [18], the following diagrams were proposed: for the requirements managing: the simulators diagram (aims at unified representing simulation models of continuous DS-processes, their ordering within

application area’s models hierarchy [18], defining topological and functional relationships between models as well), the application area diagram (contains a UML-notated description of researched DS, as well as their properties and features); for the system analysis: the advanced activity diagram (represents messages interchange and informational coupling among modelling and service DSE-software components through the special classification symbol), the timing diagram (used for specification of researched DS’s behaviour in time, documentation of time parameters of simulation study proceeding as well); for the projecting: the advanced deployment diagram (is provided by UML-notion of coupling simulation services and their components with groups of distributed DSE’s resources); for the realization: the advanced class diagram (represents object-oriented organizational structure of modelling components realization), the sequence diagram (contains DSE’s resources as participants and defines use of simulation services by solving the simulation tasks), the data diagram (aims at representing the structure of informational layer of SimGrid infrastructure).

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

44

6. CONCLUSION This article gave a short introduction into the

motivation and the current implementation of Grid infrastructures for solving tasks of industrial simulation. Although research groups all over the world are working towards development of Grid-environments that would be available to groups of simulation scientists independent of their local environment of computers, programs, data and information resources, considerable additional research and development for based on Grid technologies Simulation Service Providing in different application areas is required. As result of this a widespread adoption by scientific organizations and research groups and a broader use of Grid technologies for solving industrial tasks are expected. This requires complex approaches development, one of them - a technologically-oriented distributed simulation environment, as well as system organization of simulation services providing an easy access to Grid infrastructure of simulators developers and engineering stuff as well - is proposed in this article. The future developments will concentrate on further development of simulation models and services for dynamic systems as well as their implementation for solving industrial tasks as components of problem-oriented Simulation and Servicing Centres for specified application areas.

7. REFERENCES

[1] P. Fishwick. Web-Based Simulation: Some Personal Observations. Proceedings of the 1996 Winter Simulation Conference (1996). pp. 772-779.

[2] P.Fishwick. Using XML for Simulation Modeling. Proceedings of the 2002 Winter Simulation Conference (2002).

[3] J. Kuljis, R. Paul. A review of web based simulation: whiter we wander? Proceedings of the 2000 Winter Simulation Conference, Orlando Florida (2000). pp. 1872-1881.

[4] T. Wiedemann. Simulation Application Service Providing (SIM-ASP). Proceedings of the 2001 Winter Simulation Conference (2001).

[5] T. Wiedemann. Next Generation Simulation environments Founded On Open Source Software and XML-Based Standard Interfaces. Proceedings of the 2002 Winter Simulation Conference (2002).

[6] F. Bry, W. Nagel, M. Schroeder. Grid-Computing. Informatik Spektrum 13 (2004). pp. 552-554.

[7] F. Berman, A. Hey, G. Fox. Grid computing: Making the global infrastructure a reality.

Wiley (2003). [8] O. Cheptsov, V. Svjatnyj, R. Hohmann. Die

Entwicklung eines Simulations- und servicezentrums für gegebenes Gegenstandsgebiet. Collected Volume of Scientific Papers of Donetsk National Technical University. SCAD-2005. Volume 93 (2005). pp. 151-158.

[9] M. T. Dove and N.H. De Leeuw. Grid computing and molecular simulations: the vision of the eMinerals project. Molecular Simulation, Vol. 31, No. 5, (2005). pp. 297–301.

[10] V. Svjatnyj., A. Masjuk, O. Smagin, H.-J. Bungartz. Interaktive Modellierung, Simulation und Prozessführung in einer parallelen problemorientierten Simulationsumgebung. F.Hülsemann u.a. (Hrsg.), Tagungsband 18. ASIM-Symposium Simulationstechnik, Erlangen (2005). pp. 742-747.

[11] A. Cheptsov. The UML-Approach Of Developing Discrete Simulators (On Instance Of Coal Industry’s Objects). Simulation and Computer Graphic, Donetsk (2008). pp. 354-358.

[12] L.P. Feldmann, V.A. Svjatnyj, V.V. Lapko, E.-D. Gilles, M. Zeitz, A. Reuter, K. Rothermel. Parallel Simulation Technology. Problems of Simulation and Computer Aided Design of Dynamic systems. Collected Volume of Scientific Papers of Donetsk National Technical University. SCAD-1999 (1999). pp. 9-19.

[13] V. Svjatnyj, K. Teplinskiy, V. Trubarov. Optimisation Problems in the Technological-Oriented Parallel Simulation Environment. F.Hülsemann u.a. (Hrsg.), Tagungsband 18. ASIM-Symposium Simulationstechnik, Erlangen (2005). pp.582-587.

[14] M. Gyimesi, F. Breitenecker. Simulation Service Providing als Web Service. F.Hülsemann u.a. (Hrsg.), Tagungsband 18. ASIM-Symposium Simulationstechnik, Erlangen (2005).

[15] O. Cheptsov, V. Svjatnyj, O. Beljaev, V. Lapko, O. Schkrebez Zur Entwicklungs-organisation des Simulations- und Service-zentrums für die Kohleindustrie. Simulationstechnik 18. Symposium in Erlangen, September (2005). pp.554-559.

[16] R. Miles, K. Hamilton. Learning UML 2.0. O’Reilly (2006). 269 p.

[17] Sommerwille Ian. Software Engineering. 6.Auflage. Addison Wesley Verlag (2001).

[18] O. Cheptsov. The system organization and basic algorithms of the Simulation- and

Alexey Cheptsov / Computing, 2007, Vol. 6, Issue 3, 38-45

45

Servicecentre for the coal industry. IEEE Proceedings International Conference “Modern Problems of Radio Engineering, Telecommunications and Computer Science TCSET’2006” (2006), pp. 205-207.

Dr.-Ing. Alexey Cheptsov holds a diploma (M.Sc.) of computer systems and networks and a diploma of economics (Dipl.-Fin.) of the Donetsk National Technical University (Ukraine). He has done a Ph.D. in automated control systems and

progressive informational technologies at the G.Pukhov’s Institute of Simulation Problems in Power Engineering, National Academy of Sciences of Ukraine. At present he works as an Assistant Professor of the Department of Computer Science of the Donetsk National Technical University. He is the leader of Distributed and Parallel Computing Group. His main research interests are Grid technology, high-performance computation, embedded systems, automotive electronic engineering.