Academia.eduAcademia.edu

Timed and Stochastic Model Checking of Petri Nets

AI-generated Abstract

This paper presents a comprehensive exploration of timed and stochastic model checking applied to Petri Nets, focusing on both real-time system verification and qualitative outcomes. It introduces mechanisms for verifying properties within Petri Net models, particularly emphasizing two system classes: hard and soft real-time systems. The research ultimately showcases the efficacy of Stochastic Weighted Networks (SWN) in yielding results comparable to Generalized Stochastic Petri Nets (GSPN) while emphasizing the importance of Quality of Service (QoS) metrics in analyzing system performance.

Key takeaways

  • A transition t k is said to be newly enabled after the firing of a transition t i from a marking M if t k is not
  • Remark In order to build the MCTA of a TPN, the number of marking classes has to be bounded, otherwise the construction of the MCG will not terminate.
  • "observation" transitions are there defined as transitions with a modified semantics (they are enabled but they never fire), a change in the semantics that is not necessary in our case, if we exclude the use of the Next operator of CSL, because an exponential transition which does not change the marking does not alter the (infinitesimal generator of the) CTMC.
  • This chapter discusses how we have exploited two CSL model checkers, PRISM and MRMC, to add CSL model checking facilities for GSPN and SWN in the GREATSPN tool.
  • We have exploited two CSL model checkers, PRISM 2 and MRMC, to add CSL model checking facilities for GSPN and SWN into the GREATSPN tool; this led to the implementation of two tools GREAT2PRISM and