Numerical vs. statistical probabilistic model checking

  • Authors:
  • Håkan L. S. Younes;Marta Kwiatkowska;Gethin Norman;David Parker

  • Affiliations:
  • Computer Science Department, Carnegie Mellon University, 15213, Pittsburgh, PA, USA;School of Computer Science, University of Birmingham, B15 2TT, Birmingham, PA, UK;School of Computer Science, University of Birmingham, B15 2TT, Birmingham, PA, UK;School of Computer Science, University of Birmingham, B15 2TT, Birmingham, PA, UK

  • Venue:
  • International Journal on Software Tools for Technology Transfer (STTT)
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Numerical analysis based on uniformisation and statistical techniques based on sampling and simulation are two distinct approaches for transient analysis of stochastic systems. We compare the two solution techniques when applied to the verification of time-bounded until formulae in the temporal stochastic logic CSL, both theoretically and through empirical evaluation on a set of case studies. Our study differs from most previous comparisons of numerical and statistical approaches in that CSL model checking is a hypothesis-testing problem rather than a parameter-estimation problem. We can therefore rely on highly efficient sequential acceptance sampling tests, which enables statistical solution techniques to quickly return a result with some uncertainty. We also propose a novel combination of the two solution techniques for verifying CSL queries with nested probabilistic operators.