• DocumentCode
    2765234
  • Title

    Applying Akaroa2 for automated simulation length control & on-line statistical error analysis of results in OPNET

  • Author

    Ul Haque, M. ; Pawlikowski, K. ; McNickle, D. ; Willig, A.

  • Author_Institution
    Comput. Sci. & Software Eng. Dept., Univ. of Canterbury, Christchurch, New Zealand
  • fYear
    2012
  • fDate
    7-9 Nov. 2012
  • Firstpage
    1
  • Lastpage
    5
  • Abstract
    Simulation is used for developing and testing different scenarios under controlled and reproducible situations. Proper handling of the initial transient effect for steady state simulations, proper selection of simulation length and proper statistical analysis of results are essential for conducting credible simulations. Akaroa2, a universal controller for quantitative discrete event simulation, has been designed to improve credibility of simulation results. It automatically handles the initial transient effect, carries out statistical analysis of results and controls simulation length by stopping the simulation when the required precision is achieved based on continuous analysis of mean values. Simulation programs can be made to run under Akaroa2 to provide statistically valid results. OPNET is a popular simulator used for carrying out telecommunication and networking related simulations, but it does not provide automated simulation length control and initial transient effect handling. We have developed an interface using OPNET co-simulation capabilities to run OPNET simulations under Akaroa2 control. It will allow the user of OPNET to control the length of simulation on basis of the accuracy of statistical results. OPNET and Akaroa2 are available for free for non-profit research at universities.
  • Keywords
    control system synthesis; discrete event simulation; educational institutions; error analysis; parallel processing; statistical analysis; Akaroa2 control; OPNET simulations; automated simulation length control; conducting credible simulations; initial transient effect; mean values continuous analysis; nonprofit research; online statistical error analysis; quantitative discrete event simulation; reproducible situations; simulation length; simulation length controls; simulation programs; steady state simulations; universal controller; universities; Analytical models; Computational modeling; Delay; Educational institutions; Simulation; Steady-state; Transient analysis; Akaroa2; Interface; OPNET; Statistical Anylasis; co-simulation; simulation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Telecommunication Networks and Applications Conference (ATNAC), 2012 Australasian
  • Conference_Location
    Brisbane, QLD
  • Print_ISBN
    978-1-4673-4408-1
  • Electronic_ISBN
    978-1-4673-4409-8
  • Type

    conf

  • DOI
    10.1109/ATNAC.2012.6398076
  • Filename
    6398076