Previous Next Methods Allocation  
4.47 Method Category "Reliability Models" (RELM)  

  4.47 Methodenkategorie "Zuverlässigkeitsmodelle" (ZUVM)

Contents  
  • 1 Identification/Definition of the Method
  • 2 Brief Characteristic of the Method
  • 3 Limits of the Methods Application
  • 4 Specification of the Methods Allocation
  • 5 Interfaces
  • 6 Further Literature
  • 1 Identification/Definition of the Method

    Method "Reliability Models" (RELM) is a category of methods, the individual applicable reliability models are described in a more detailed way in RELM - Reliability Models (in Annex 1), by listing the selection criteria. A special method is only defined within the scope of the operationalization.

    /Goel, 1985/ can be used as a general description.

    2 Brief Characteristic of the Method

    Objective and Purpose

    A Reliability Model is applied for the identification, the compression and the proof of reliability requirements. Based on the user-oriented requirements and environment, the system is to be represented by the model either complete or adaptive.

    The Reliability Models is not only to inform about achieving the user-oriented quality goals but also about criteria in connection with these goals, as well as about the interim goals to be achieved (reliability increase) and the influence of technical changes.

    Operational Sequence

    The principle of the method is to develop a model that makes it possible to obtain information about the functions of a software/a system and the corresponding reliability state.

    In order to measure the reliability state, different reliability characteristics are applied. The characteristics most frequently applied are:

  • mean time between failure (MTBF)
  • failure rate observed in a certain time interval
  • estimated number of faults remaining
  • The characteristic "Mean Time between Failures" is based on measuring the time interval between the occurrence of failure n and the occurrence of failure n+1. In the case where an acceptable value has been reached for the time interval, both development and quality assurance can be finished.

    The characteristic "Failure Rate" counts or estimates the number of failures during the predefined interval. If the cumulative number of occurred failures has reached an acceptable (low) number during the time interval, both development and quality assurance can be finished.

    The characteristic "Estimated Number of Faults Remaining" estimates the number of faults remaining in the program. If this number is acceptable, both development and quality assurance can be finished.

    Realization and application of a reliability model are closely connected with the measuring process for the reliability state.

    3 Limits of the Methods Application

    The Method RELM is to be applied in the case chen the reliability requirements are high.

    4 Specification of the Methods Allocation

    No. Activity Description
    4.1 SD1.3 - Definition of Criticality and Quality Requirements Method RELM is applied for the definition of requirements for the reliability of system functions. It completely covers subproduct User Requirements.Quality Requirements, but only with regard to the reliability.
    4.2 SD3.4 - Definition of Requirements for the Quality of the SW/HS Unit Method RELM is applied to define the requirements for the reliability of system functions. It completely covers subproduct 5.x.4.3.Technical Requirements.Technical Requirements for Other Quality Characteristics, but only with regard to reliability.

    5 Interfaces

    - not applicable -

    6 Further Literature

    /Angus, 1984/ The Application of Software Reliability and Maintainability
    /Goel, 1985/ Software Reliability Models: Assumptions, Limitations, and Applicability
    /IEEE 982.2/ IEEE-STD 982.2-1988 - IEEE Guide for the Use of IEEE Standard Dictionary of Measures to Produce Reliable Software, 1988
    /Lapadula, 1975/ Engineering of Quality Software Systems, Volume VIII: Software Reliability Modelling and Measurement Technique
    /Miller, 1986/,
    /MIL-HDBK 189, 1981/ Reliability Growth Management
    /MIL-STD 756B, 1982/ Reliability, Modelling and Prediction
    /Srivastava, 1986/ The Use of Software Reliability Models in the Analysis of Operational Systems
    /Sukert, 1977/ An Investigation of Software Reliability Models
    /Yamada, 1985/ Software Reliability Growth Modelling: Models and Applications

    Previous Next GDPA Online Last Updated 01.Jan.2002 Updated by Webmaster Last Revised 01.Jan.2002 Revised by Webmaster