• DocumentCode
    723494
  • Title

    Modelling burst errors in industrial networks

  • Author

    Platschek, Andreas ; Thiemann, Bernd ; Zeilinger, Heimo

  • Author_Institution
    Inst. for Comput. Technol., Vienna Univ. of Technol., Vienna, Austria
  • fYear
    2015
  • fDate
    27-29 May 2015
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    When argumenting the safety of an industrial system, the validity of the calculation of the residual error rate depends on multiple factors. Examples are a correct and thorough analysis of hazards, a correct estimation of occurrence and on the risk the identified hazards pose. These properties of the analysis have then to be compiled into a sound error model that reflects reality, and can then be used as basis for calculating the residual error rate of the system. One error model for industrial communication networks is given in IEC 61784-3 [1, B.3]. In this paper, this commonly used error model for industrial communication networks is analyzed and its shortcomings are pin-pointed, leading to the question, whether the current model should be scrutinized, not only its parameters. Based on this analysis a new approach on an error model is proposed. This new model has the advantages of more accurately modelling burst errors, and using input parameters that give less reason for discussion as they are widely used in electro magnetic compatibility (EMC) tests for hardware.
  • Keywords
    computer networks; electromagnetic compatibility; EMC test; IEC 61784-3; burst error modelling; electromagnetic compatibility; industrial communication network; industrial system safety; residual error rate; sound error model; Adaptation models; Error analysis; IEC standards; Industrial communication; Mathematical model; Probability; Safety;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Factory Communication Systems (WFCS), 2015 IEEE World Conference on
  • Conference_Location
    Palma de Mallorca
  • Type

    conf

  • DOI
    10.1109/WFCS.2015.7160587
  • Filename
    7160587