Scribd
Upload
75 views5 pages

TimeNET QEST2017

This document describes new features in version 4.4 of the software tool TimeNET, which supports modeling and performance evaluation of stochastic models including Petri nets and Markov chains. Key updates include algorithms for efficient rare-event simulation of Petri nets using techniques like splitting, importance sampling integration, and a new hybrid simulation/analysis algorithm. The tool now also supports modeling and analysis of Markov chains, discrete-time stochastic automata, and UML models with stochastic elements. Various improvements were made to the graphical user interface, simulation methods, and software development processes.

Uploaded by

Mostafa J Alneamy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
75 views5 pages

TimeNET QEST2017

This document describes new features in version 4.4 of the software tool TimeNET, which supports modeling and performance evaluation of stochastic models including Petri nets and Markov chains. Key updates include algorithms for efficient rare-event simulation of Petri nets using techniques like splitting, importance sampling integration, and a new hybrid simulation/analysis algorithm. The tool now also supports modeling and analysis of Markov chains, discrete-time stochastic automata, and UML models with stochastic elements. Various improvements were made to the graphical user interface, simulation methods, and software development processes.

Uploaded by

Mostafa J Alneamy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319039466

Modelling and Performance Evaluation with TimeNET 4.4

Conference Paper · August 2017


DOI: 10.1007/978-3-319-66335-7_19

CITATIONS READS
29 381

1 author:

Armin Zimmermann
Technische Universität Ilmenau
156 PUBLICATIONS   1,777 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Model-Based Reliability Evaluation View project

TimeNET View project

All content following this page was uploaded by Armin Zimmermann on 17 August 2017.

The user has requested enhancement of the downloaded file.


Modelling and Performance Evaluation
With TimeNET 4.4

Armin Zimmermann

Department of Computer Science and Automation


Systems and Software Engineering Group, TU Ilmenau, Germany
armin.zimmermann@tu-ilmenau.de
WWW home page: http://timenet.tu-ilmenau.de

Abstract. The paper presents the current status of the software tool
TimeNET. It supports modeling and performance evaluation of stochas-
tic models, including extended deterministic and stochastic Petri nets,
colored stochastic Petri nets, and Markov chains as well as UML ex-
tensions. Among its main characteristics are simulation and analysis
modules for stationary and transient evaluation of Petri nets includ-
ing non-exponentially distributed delays, as well as a simulation module
for complex colored models. Recent enhancements include algorithms for
the efficient rare-event simulation of Petri nets, a new multi-trajectory
hybrid simulation/analysis algorithm, and a net class for Markov chains.

Keywords: Modeling tool, TimeNET, stochastic Petri nets, colored


Petri nets, performance evaluation

1 Introduction
TimeNET is a software tool for the modeling and performability evaluation with
several variants of stochastic Petri nets including GSPNs, eDSPNs, and colored
SCPNs (for definitions see e.g. [8]). In comparison to other related tools such as
GreatSPN [1], SPNP [4], Möbius [3] and CPN Tools [5], it supports evaluation of
models combining exponential and deterministic as well as more general non-ex-
ponentially distributed firing delays. Numerical analysis and simulation methods
both for transient and steady-state solution have been implemented as well as
structural analysis modules. Moreover, TimeNET supports colored stochastic
Petri nets as well as rare-event simulation algorithms for these model classes.
The token game can be run interactively or automatically to validate and test
eDSPN and SCPN models.
The software architecture contains a Java graphical user interface, shell scripts
controlling analysis processes, and evaluation algorithms implemented mainly in
C++ running as background processes. The tool runs in 32 and 64 Bit Linux
and Windows environments. It is available free of charge for non-commercial use
from http://timenet.tu-ilmenau.de/. Successful applications reported in the
literature include communication systems, reliability evaluation, manufacturing
and transportation, and business as well as logistics processes. Numerous papers
including application examples are listed in the tool’s web page.
2

2 New Features in TimeNET 4.4


This paper presents changes in TimeNET since the previous tool description [9],
which covered version 4.1 in 2012. More in-depth coverage of history and tool
architecture can be found in [9, 8].
Among the various changes in the tool since 2012, the scientifically most
relevant extensions cover rare-event simulation methods motivated by reliability
applications. Such examples will otherwise lead to unacceptably long run times
because of the number of events to be simulated until enough samples of interest
are generated.
An example application model is shown in Figure 1, describing a sample
network architecture of the Avionic Full-Duplex Ethernet used in modern air-
craft [11]. Reliability of such systems is a major concern, and a model-based
analysis can show that the required end-to-end message delays are achievable
for a certain setup. This is a typical example of industrial systems in which nu-
merical analysis is impossible because of the large state space and concurrent
non-Markovian activities, while standard simulation would need exceedingly high
run times to compute the results with acceptable statistical accuracy.
While rare-event simulation is a well-known technique for the efficient evalu-
ation of highly reliable systems, the available algorithms require significant back-
ground knowledge or apply to quite restricted model classes only. Our goal is to
make such methods available for tool users, aiming at semi-automatic algorithm
configuration using the available model information.
TimeNET implements a variant of the splitting technique RESTART, which
is now extended by automatically deriving an estimation of the result via perfor-

Fig. 1. AFDX network model


3

mance bounds to calculate splitting factors, and an online distance estimation


using structural properties analyzed via a linear programming problem [12]. This
avoids the user-defined importance function otherwise necessary. In addition to
that, an automated importance sampling method for rare-event SPN simulation
proposed by Daniël Reijsbergen [6] is currently being integrated in the tool.
A new multi-trajectory simulation algorithm for eDSPNs [10] has been im-
plemented recently. It combines elements of simulation and numerical analysis
for the first time such that the behavior of the performance evaluation method
follows either method just depending on the amount of trajectories (state par-
ticles) being stored and followed. This allows a mixed approach, avoiding the
pitfalls of simulation (rare events) and numerical analysis (large state spaces).
Another addition is a simulation method for eDSPN models with incom-
pletely known initial state. Automated and distributed optimization of eDSPN
and SCPN models [2] have been implemented as a separate add-on. The initial
transient phase of steady-state SCPN simulations is now detected and deleted.
Modern random number generators for simulations (Mersenne twister) are used
instead of standard library implementations now.
Model extensions include a new model class stochastic automata that al-
lows to specify discrete-time and continuous-time Markov chains for teaching
purposes. A standard stationary solution has been implemented. UML state
charts extended with stochastic elements based on the MARTE profile and en-
ergy usage stereotypes have been added as another model class [7]. Such models
can be edited and translated into eDSPNs models for the analysis of embedded
systems. Model parameters (or definitions) are now unified without a numeric
type and may depend on each other in most model classes. Reward definitions
for performance measures have been unified for eDSPNs and SCPNS, and ex-
tended by transition throughputs (impulse rewards). Color-dependent SCPN
model parts were extended and streamlined, including performance measures,
arc expressions, and firing delays.
GUI and user interaction have been improved with several details, in-
cluding a graphical visualization of the reachability graph, and adaptive mouse
pointer appearance when adding objects.
User support outside the tool itself is undergoing a major update in 2017:
A new integrated support web page will be made available until summer 2017. It
contains installation and background information, example models, FAQs, daily
and by-version downloads, and user feedback for bug reports etc.
Efficiency and quality of the internal software development process are
now supported by a build and test server that automatically and regularly checks
out the complete source code on different supported systems, compiles the tool,
tests it, and makes new builds available for download. The tool is now available
as a native 64 Bit application.
The author wishes to thank the numerous students who have contributed to
the development of TimeNET over the years.
4

References
1. Bernardi, S., Bertoncello, C., Donatelli, S., Franceschinis, G., Gaeta, G., Gribaudo,
M., Horvàth, A.: GreatSPN in the new millenium. Research Report 760/2001,
Universität Dortmund (Germany) (Sep 2001), tools of Aachen 2001, Int. Multiconf.
on Measurement, Modelling and Evaluation of Computer-Communication Systems,
pages 17–23.
2. Bodenstein, C., Zimmermann, A.: TimeNET optimization environment - batch
simulation and heuristic optimization of SCPNs with TimeNET 4.2. In: 8th Int.
Conf. on Performance Evaluation Methodologies and Tools (VALUETOOLS 2014).
Bratislava, Slovakia (Dec 2014)
3. Courtney, T., Gaonkar, S., Keefe, K., Rozier, E., Sanders, W.: Möbius 2.3: An
extensible tool for dependability, security, and performance evaluation of large
and complex system models. In: IEEE/IFIP Int. Conf. on Dependable Systems
Networks. pp. 353–358 (2009)
4. Hirel, C., Tuffin, B., Trivedi, K.S.: SPNP: Stochastic Petri nets. Version 6.0. In:
Computer Performance Evaluation, Modelling Techniques and Tools — 11th Int.
Conf., TOOLS 2000. Lecture Notes in Computer Science, vol. 1786, pp. 354–357.
Springer Verlag, Schaumburg, IL, USA (2000)
5. Jensen, K., Kristensen, K.L., Wells, L.: Coloured Petri nets and CPN tools for
modelling and validation of concurrent systems. Int. Journal on Software Tools for
Technology Transfer (STTT) 9(3–4), 213–254 (2007)
6. Reijsbergen, D., Boer, P.T., Scheinhardt, W., Haverkort, B.: Automated rare
event simulation for stochastic Petri nets. In: Joshi, K., Siegle, M., Stoelinga, M.,
D’Argenio, P.R. (eds.) Quantitative Evaluation of Systems, Lecture Notes in Com-
puter Science, vol. 8054, pp. 372–388. Springer Berlin Heidelberg (2013)
7. Shorin, D., Zimmermann, A.: Extending the software tool TimeNET by power con-
sumption estimation of UML MARTE models. In: Proc. 4th Int. Conf. on Simula-
tion and Modeling Methodologies, Technologies and Applications (SIMULTECH
2014). pp. 83–91. Vienna, Austria (Aug 2014)
8. Zimmermann, A.: Stochastic Discrete Event Systems. Springer, Berlin Heidelberg
New York (2007)
9. Zimmermann, A.: Modeling and evaluation of stochastic Petri nets with TimeNET
4.1. In: Proc. 6th Int. Conf on Performance Evaluation Methodologies and Tools
(VALUETOOLS). pp. 54–63. Corse, France (2012)
10. Zimmermann, A., Hotz, T., Canabal Lavista, A.: A hybrid multi-trajectory simu-
lation algorithm for the performance evaluation of stochastic Petri nets. In: Quan-
titative Evaluation of Systems (QEST 2017) (2017), (accepted for publication)
11. Zimmermann, A., Jäger, S., Geyer, F.: Towards reliability evaluation of AFDX
avionic communication systems with rare-event simulation. In: Proc. Probabilistic
Safety Assessment & Management Conference 2014 (PSAM 12). pp. 1–12. Hon-
olulu, Hawaii, USA (Jun 2014)
12. Zimmermann, A., Maciel, P.: Importance function derivation for RESTART sim-
ulations of Petri nets. In: 9th Int. Workshop on Rare Event Simulation (RESIM
2012). pp. 8–15. Trondheim, Norway (Jun 2012)

View publication stats

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy