SWZ  >  
Research Projects  >  
DeSim: Distributed Architectures and Concepts for the Simulation of Systems of Systems

DeSim: Distributed Architectures and Concepts for the Simulation of Systems of Systems

The demands on distributed simulation environments is ever  increasing. It will become more and more essential to simulate and analyze not the behavior of a single system but that of complex Systems of Systems (SoS).
Traditional methods focus on speeding up the simulation itself (e.g. through parallelization) and mostly lack the capability to execute and especially create and design the models correspondent to these SoS. Consider e.g. a complex value creation chain with existing models for each of its aspects. It might not be desirable to bring all this information together in one system, e.g. because of security/confidentiality concerns of the involved parties.
To tackle this problem, Multi-Agent-Based Simulation (MABS) seems to be a most promising and quickly evolving key concept. It allows for microscopic modelling of actors, their goals, preferences capabilities and interaction between them, which is rather difficult, if not impossible, using traditional physical or cellular models.
Current agent-oriented models, e.g. in the field of traffic simulation, already support interaction, coordination and cooperation models, however, they are very restrictive in that and do not support distributing the models to multiple runtime environments.
The scope of this project comprises finding basic architectures, concepts and requirements for decentral modelling and execution of simulations of SoS.
Based on this, reasonable and suitable methods, models, protocols and utilities from the field of multi-agent systems are to be identified and - if possible - implemented in an existing framework (the Multi-Agent System simulation platform [MASSim], which is used for the annual Multi-Agent Programming Competition) and/or a completely new framework (the Parallel Multi-Agent simulation platform [PMAS]).

Involved Scientists

Prof. Dr. Jürgen Dix

Research Group Computational Intelligence
Department of Informatics
Clausthal University of Technology

Room 116, Am Regenbogen 15, 38678 Clausthal-Zellerfeld
E-Mail: juergen.dix@tu-clausthal.de
Phone: +49 5323 72-7181
Fax: +49 5323 72-7189

Prof. Dr. Jörg Müller

Business Information Technology Unit
Department of Informatics
Clausthal University of Technology

Room 201, Julius-Albert-Straße 4, 38678 Clausthal-Zellerfeld
E-Mail: joerg.mueller@tu-clausthal.de
Phone: +49 5323 72-7141
Fax: +49 5323 72-7199

Publication list

2016

2015

2014

2013