Static detection of zeno runs in UPPAAL networks based on synchronization matrices and two data-variable heuristics

  • Authors:
  • Jonas Rinast;Sibylle Schupp

  • Affiliations:
  • Institute for Software Systems, Hamburg University of Technology, Hamburg, Germany;Institute for Software Systems, Hamburg University of Technology, Hamburg, Germany

  • Venue:
  • FORMATS'12 Proceedings of the 10th international conference on Formal Modeling and Analysis of Timed Systems
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper addresses Zeno runs, i.e., transition sequences that can execute arbitrarily fast, in the context of model checking with the UPPAAL tool. Zeno runs conflict with real-world experience where execution always takes time and they may introduce timelocks into the models. We enhance previous work on static detection of Zeno runs by extending synchronization exploitation using a synchronization matrix that not only ensures valid synchronization partners exist but also that their number is correct. Additionally, we introduce two data-variable heuristics into the analysis as in most models data-variable constraints prevent certain Zeno runs. The analysis is implemented in a tool called ZenoTool and empirically evaluated using 13 benchmarks. The evaluation shows that our analysis is more accurate in 3 cases and never less accurate than the analysis results of previous work and that performance and memory overhead are at the same time very low.