CS 780 Spring 2007
Title: Debugging
Simulation Models
Instructor: Peter
Kemper
Prerequisites:
A first course on
simulation is recommended
(or permission of the
instructor)
Topic:
Simulation modeling is
widely applied in practice due to its few restrictions.
Many tools support
simulation to make a model-based design and evaluation of systems convenient
and productive.
The crux in simulation
is not to compute numbers but to obtain reasonable accurate ones with respect
to the real system that is modeled.
In the broader context
of verification and validation of simulation models, this research seminar
focuses on techniques and tool support to trace bugs in the code of discrete
event simulation models.
It will cover a
variety of techniques including model checking, tracking of performance
measures and visualization techniques that are based on simulation
traces.
The research seminar
covers current topics in verification and validation techniques for simulation
modeling, with an emphasis on debugging techniques based on trace analysis.
Topics include run-time verification, model checking, statistics, evaluation of
existing techniques and tools.
Students read and present technical papers from relevant journals and conference proceedings. Synthesis and understanding of materials is demonstrated by a required research project, which includes preparation of a proposal, performing the work, writing a report in the form of a research paper and presentation of results. Given the practical relevance of the overall topic, the individual research topics are likely to involve derivation, refinement, evaluation or implementation of applied analysis techniques, evaluation of simulation case studies and existing tools. Students can work in pairs on a project in any of the areas that the course will cover.