Selected Publications
2000
- Dobán, O.; Pataricza, A.:
Cost Estimation Driven Software Project Management
in UML
Proceedings of design and diagnostics of electronic circuits and systems workshop,
Elena Gramatova, Hans Manhaeve and Adam Pawlak (editors),
Institute of Informatics, Slovak Academy of Science, pp. 34, 2000
While in the former decades the design for dependability issue was confined
to hard mission critical systems, the widespread use of online services raises
the same question in an increasingly wide spectrum of applications. However,
there exist basic differences between the design of mission critical
applications and on-line services. Mission critical applications can
tolerate a high degree of redundancy and the related cost overhead compared
to a pure functional solution. However everyday's applications require only
a high availability of the applications as an add-on to the quality of the
system, but do not allow for a significant overhead.
- A. Pataricza, Gy. Csertán, Gy. Román, V. Keresztély,
J. Sziray:
UML Based Control of Industrial Processes
Proceedings INES-2000, the 2000 IEEE International Conference on Intelligent
Engineering Systems.
This paper presents a new methodology for designing control of industrial
processes. The methodology has been developed jointly by the Department of
Measurement and Information Systems at the Technical University of Budapest,
the Department of the Information Technology at the Széchenyi College,
and the Department of Information Systems at the Budapest University of Economics.
The objective of the system is the development of automated synthesis of mathematical
models of production optimization problems, where the requirements specification
is described by means of state of the art object-oriented CASE tools.
- Huszerl G., Majzik I.:
Quantitative Analysis of Dependability Critical
Systems Based on UML Statechart Models
Proc. HASE'2000, the Fifth IEEE International High-Assurance Systems Engineering
Symposium
pp 83-92, 2000
The paper introduces a method which allows quantitative performance and dependability
analysis of systems modeled by using UML statechart diagrams. The analysis
is performed by transforming the UML model to Stochastic Reward Nets (SRN).
A large subset of statechart model elements is supported including event processing,
state hierarchy and transition priorities. The transformation is presented
by a set of SRN design patterns. Performance measures can be directly derived
using SRN tools, while dependability analysis requires explicit modeling of
erroneous states and faulty behavior.
- Huszerl G., K. Kosmidis, M. Dal Cin, Majzik I. and Pataricza A.:
Quantitative Analysis of UML Statechart Models of
Dependable Systems
Submitted to The Computer Journal, 2000
The paper introduces a method which allows quantitative performance and
dependability analysis of systems modeled by using UML statechart diagrams.
The analysis is performed by transforming the UML model to Stochastic Reward
Nets (SRN). A large subset of statechart model elements is supported including
event processing, state hierarchy and transition priorities. The transformation
is presented by a set of SRN patterns. Performance measures can be directly
derived using SRN tools, while dependability analysis requires explicit
modeling of erroneous states and faulty behavior. Keywords: UML statecharts,
dependability analysis, performance analysis.
- Huszerl G. and K. Kosmidis:
UML - Extensions for Quantitative Analysis
In Proc. of UML 2000 Workshop: Dynamic Behaviour in UML Models: Semantic Questions
, LMU-München, Institut für Informatik, Bericht 0006 York, UK, October,
2000
The Unified Modeling Language (UML) finds more and more applications. It is
not only used for software development but also for modelling systems with
dynamic behavior (e.g. Flexible Manufacturing Systems (FMS), or Business Process
Modeling). While the static diagrams of UML were changed marginal in the latest
versions of UML, the dynamic diagrams need still to be improved. For modelling
truly the dynamic behavior by these diagrams a concept of time is needed.
Therefore, we introduce in this paper an approach to extend UML with time
as stochastically variable. To evaluate the enhanced models and to drive a
numerical analysis we use the Petri Net analysis tool PANDA. This tool works
also on some stochastic extensions of the Petri Nets (GSPN's), widely used
for performance evaluation and numerical analysis. The evaluation of these
models is based on exploring and solving the underlying Markov chains.
- A. Petri:
Combination of IDDQ Testing and High Level ATPG
In: Proceedings of the DDECS 2000 Workshop, April 2000, Smolenice, Slovak Republic,
pp. 63-66.
Test generation for today's complex digital circuits is an extremely computation
intensive task. The search space of ATPG can be reduced by starting from higher
level circuit descriptions. The integration of alternate testing methodology
- IDDQ testing - is suggested for increasing the efficiency of a high level
VHDL based test generator.
- Dániel Varró, Gergely Varró and András Pataricza:
Designing the Automatic Transformation of Visual
Languages
In Hartmut Ehrig and Gabriele Taentzer (eds.): GRATRA 2000 Joint APPLIGRAPH
and GETGRATS Workshop on Graph Transformation Systems, pp. 14-21., Technical
University of Berlin, Germany, March 2000
BibTex
Abstract: The process of developing dependable,
safety-critical systems controlled by computers
requires a formal verification of conceptual and
architectural choices by using different mathematical
tools. According to a novel approach of IT system
design, these input models to formal mathematical
analysis are transformed automatically from the system
model. Up to now, the design and implementation of
such transformations was rather ad hoc missing any
formal descriptions and methods. In this paper we
present our efforts towards a model transformation
system based on a powerful integration of graph
transformation, planner algorithms and deductive
databases in order to obtain an automatically
generated, provenly correct and complete
transformation code.
- Dániel Varró, Gergely Varró and András Pataricza:
Visual Graph Transformation in System Verification
In: Elena Gramatova, Hans Manhaeve and Adam Pawlak (eds.): DDECS 2000 Design
and Diagnostics of Electronic Circuits and Systems, pp. 137-141, Institute of
Informatics, Slovak Academy of Sciences, Bratislava Slovakia, April 2000.
BibTex
Abstract: The use of formal verification methods is
essential in the design process of dependable computer
controlled systems. A complex environment should
support the semi-formal specification as well as the
formal verification of the desired system. The
efficiency of applying these formal methods will be
highly increased if the underlying mathematical
background is hidden from the designer. In such an
integrated system effective techniques are needed to
transform the system model to different sort of
mathematical models supporting the assessment of
system characteristics. The current paper introduces
our research results towards a general-purpose model
transformation engine. This approach results in
yielding a provenly correct and complete
transformation code by combining the powerful
techniques of graph transformation, planner algorithms
and deductive databases. Keywords: formal
verification, graph transformation, visual languages,
planner algorithms, deductive databases.
- Dániel Varró and András Pataricza:
Mathematical Model Transformation for System Verification
Technical report of the Budapest University of Technology and Economics, Department
of Measurement and Information Systems, June 2000.
Abstract: The design process of complex, dependable
systems requires a precise verification of design
decisions during the system modelling phase using
formal methods. For that reason, the mathematical
models of various formal verification tools are
planned to be automatically derived from the system
model usually constructed from UML-diagrams. In the
paper, a general framework for an automated model
transformation system is introduced providing a
uniform formal description method of such
transformations by applying the powerful computational
paradigm of graph transformation.
Keywords:
system verification, model transformation, graph
transformation, UML.
- Dániel Varró, Gergely Varró and András Pataricza:
Towards an XMI-Based Model Interchange Format for
Graph Transformation Systems
Technical report of the Budapest University of Technology and Economics, Department
of Measurement and Information Systems, September 2000
Abstract: Tools developed to demonstrate the practical
applicability of the mainly theoretical foundations of graph
transformation systems are of immense importance. Therefore,
the interaction (moreover, integration) of these tools is a
major challenge for the graph transformation (GraTra)
community in order to increase the efficiency of international
(academic or industrial) research. A first step towards
integration is a standardized, common model interchange format
providing a uniform graph description and rule representation
that is capable of handling the most fundamental concepts of
graph transformation. A potential candidate is the novel
standard the web, the Extensible Markup Language (XML), which
allows the interchange of models in a distributed environment
(i.e. Internet). In this report, we demonstrate on several
examples that a common, XML-based GraTra model interchange
format should be constructed from a metamodel following the
concepts of the Meta Object Facility standard; XMI (XML
Metadata Interchange) is the most suitable XML-based language
for such a format; the proposed XMI format can be derived
automatically from the corresponding GraTra metamodel.
- Dániel Varró, Péter Domokos and András Pataricza:
UML Specification of Model Transformation Systems
Technical report of the Budapest University of Technology and Economics,
Department of Measurement and Information Systems, October 2000
Abstract: The design of complex, dependable systems requires a precise formal
verification of design decisions during the system modelling phase. For that
reason, the mathematical models of various formal verification tools are planned
to be automatically derived from the system model usually described by UML-diagrams.
In the current paper, a general framework for an automated model transformation
system is introduced providing a uniform formal description method of such
transformations by applying the powerful computational paradigm of graph transformation.
Model transformation rules are constructed in a modular way using a visual
UML notation as representation in order to provide a closer correspondence
with industrial techniques.
Keywords: system verification, model transformation, graph transformation,
transformation unit, UML.
- Tamás Bartha, Endre Selényi:
On Classification Heuristics of Probabilistic SystemLevel
Fault Diagnostic Algorithms
Accepted for the DAPSYS2000 Austrian-Hungarian Workshop, Balatonfüred,
Hungary, Sep 2000
Abstract: Systemlevel fault diagnosis of massively parallel computers requires
efficient algorithms, handling a many processing elements in a heterogeneous
environment. Probabilistic fault diagnosis is an approach to make the diagnostic
problem both easier to solve and more generally applicable. The price to pay
for these advantages is that the diagnostic result is no longer guaranteed
to be correct and complete in every fault situation. In an earlier paper the
authors presented a novel methodology, called local information diagnosis,
and applied it to create a family of probabilistic diagnostic algorithms.
This paper examines the identification of faultfree and faulty units in detail
by defining three heuristic methods of fault classification and comparing
the diagnostic accuracy provided by these heuristics using measurement results.
Keywords: multiprocessor systems, systemlevel fault diagnosis, probabilistic
algorithms
- Tamás Bartha, Piero Maestrini:
A Backward Error Recovery Scheme for the APEmille
Parallel Computer
Accepted for the EWDC-11 Workshop, Budapest, Hungary, May 2000
Abstract: The APE supercomputers are number-crunching engines aimed at solving
complex scientific problems. The typical duration of a computation ranges
from several days to several weeks. Due to the large number of components
built into an APEmille machine, the expected MTTF without fault tolerance
may fall into the same range. Thus, a long-running computation could be invalidated
by a system failure with high probability, and there would be no guarantee
that the results of a successfully terminated program are correct. To improve
on the situation the designers of decided to incorporate fault tolerance into
the system. They chose the error removel based approach, composed of error
detection, system-level fault diagnosis, system repair, and backward error
recovery. This paper presents a failure model of the APEmille components and
develops a comprehensive recovery scheme for the whole computer, including
even the reliable storage that is used to record the recovery information.
Keywords: fault-tolerant systems, computer architectures, parallel processors,
error recovery, checkpointing, distributed databases
- Tamás Bartha:
Utilizing Backward Error Recovery To Achieve
Fault Tolerance in a SIMD Supercomputer
Accepted for IFAC SafeProcess 2000 Conference
Abstract: The APE supercomputers are number-crunching engines aimed at solving
complex scientific problems. The typical duration of a computation ranges
from several days to several weeks. Due to the large number of components
built into an APEmille machine, the expected MTTF without fault tolerance
may fall into the same range. Thus, a long-running computation could be invalidated
by a system failure with high probability, and there would be no guarantee
that the results of a successfully terminated program are correct. To improve
on the situation the designers of decided to incorporate fault tolerance into
the system. They chose the error removel based approach, composed of error
detection, system-level fault diagnosis, system repair, and backward error
recovery. This paper presents a failure model of the APEmille components and
develops a comprehensive recovery scheme for the whole computer, including
even the reliable storage that is used to record the recovery information.