A probabilistic extension of UML statecharts: specification and verification

    Research output: Book/ReportReportOther research output

    40 Citations (Scopus)
    104 Downloads (Pure)

    Abstract

    This paper is the extended technical report that corresponds to a published paper [14]. This paper introduces means to specify system randomness within UML statecharts, and to verify probabilistic temporal properties over such enhanced statecharts which we call probabilistic UML statecharts. To achieve this, we develop a general recipe to extend a statechart semantics with discrete probability distributions, resulting in Markov decision processes as semantic models. We apply this recipe to the requirements-level UML semantics of [8]. Properties of interest for probabilistic statecharts are expressed in PCTL, a probabilistic variant of CTL for processes that exhibit both non-determinism and probabilities. Verification is performed using the model checker Prism. A model checking example shows the feasibility of the suggested approach.
    Original languageUndefined
    Place of PublicationEnschede
    PublisherCentre for Telematics and Information Technology (CTIT)
    Number of pages25
    Publication statusPublished - Sept 2002

    Publication series

    NameCTIT technical report series
    PublisherUniversity of Twente, Centre for Telematics and Information Technology (CTIT)
    No.TR-CTIT-02-31

    Keywords

    • UML statecharts
    • Model Checking
    • EWI-1300
    • Semantics
    • IR-56024
    • probabilities
    • Markov Decision Processes

    Cite this