TM8105 - Advanced Discrete Event Simulation Methodology

About

Lessons are not given in the academic year 2018/2019

Course content

The course is taught every second year, next time spring 2020. The course is about simulation methods, like process oriented simulation, Markov-simulation, trace-driven simulation. Objects, mechanisms and primitives in discrete event simulation. Development of simulators based on the previously mentioned issues. (Various relevant tools/languages will be presented, discussed and used in exercises.) Planning of experiments with emphasis on control of the uncertainty (error) in the results. Statistical analysis of simulation results and presentation of results. As a part of this, techniques like replication, sectioning (batch mean), bootstrapping, jackknifing. Variance reducing techniques like control variables, stratified sampling, restart/splitting, importance sampling.

Learning outcome

A. Knowledge:
1) Overview over methods for discrete events simulation, as well as know of their strengths and weaknesses.
2) Knowledge of some commonly used simulators / simulation tools.
3) Knowledge of the basic elements of a discrete event simulator, specifically the handling of eventlists.
4) Knowledge of techniques to reduce variance and shorten the simulation times. Understanding the theoretical basis for these and the challenges of applying them.
5) Firm knowledge of the planning of simulation studies and analysis of simulation results
keeping control of the statistical uncertainty.

B. Skills:
1) Be able to develop simulators for performance and reliability studies of ICT systems. As a minimum, object-oriented simulation (prior knowledge required) and Markov simulation should be mastered.
2) Set up and carry out simulation studies.
3) Analyze simulation results applying adequate statistical methods.
4) Present the results from studies of complex systems with many parameters.

C. General competence:
1) Have a firm understanding of the simulation with discrete events as an evaluation method in a broad context.
2) Advanced knowledge of analysis and presentation of stochastic / probabilistic data.

Learning methods and activities

Colloquia/interactive lectures, where it is expected that the students have familiarized themselves with the topic beforehand. Optional exercises. If there are more than 4 candidates a written exam will be considered. If there is a re-sit examination, the examination form may be changed from written to oral.

The grading rule is pass/fail. The minimum passing grade is 70/100 points (70%).

Required previous knowledge

None

Course materials

Announced at the beginning of the term. Excerpts from textbooks, which may be supplemented by journal and conference papers, etc. Manuals for simulation tools for exercises.

Timetable

Detailed timetable

Examination

  • * The location (room) for a written examination is published 3 days before examination date.
If more than one room is listed, you will find your room at Studentweb.