Monitor-Based Statistical Model Checking for Weighted Metric Temporal Logic

Peter Bulychev (1) , Alexandre David (1) , Kim Guldstrand Larsen (2) , Axel Legay (3) , Guangyuan Li (4) , Danny Bogsten Poulsen (1) , Amélie Stainer (5)

1 CISS - Center for Indlejrede Software Systemer
2 AAU - Aalborg University [Denmark]
3 DISTRIBCOM - Distributed and Iterative Algorithms for the Management of Telecommunications Systems
4 Institute of Computational Mathematics and Scientific/Engineering Computing
5 VERTECS - Verification models and techniques applied to testing and control of reactive systems

We present a novel approach and implementation for analysing weighted timed automata (WTA) with respect to the weighted metric temporal logic (WMTL≤). Based on a stochastic semantics of WTAs, we apply statistical model checking (SMC) to estimate and test probabilities of satisfaction with desired levels of confidence. Our approach consists in generation of deterministic monitors for formulas in WMTL≤, allowing for efficient SMC by run-time evaluation of a given formula. By necessity, the deterministic observers are in general approximate (over- or under-approximations), but are most often exact and experimentally tight. The technique is implemented in the new tool Casaal that we seamlessly connect to Uppaal-smc in a tool chain. We demonstrate the applicability of our technique and the efficiency of our implementation through a number of case-studies.

Domaines: Informatique et langage [cs.CL]

Peter Bulychev, Alexandre David, Kim Guldstrand Larsen, Axel Legay, Guangyuan Li, et al.. Monitor-Based Statistical Model Checking for Weighted Metric Temporal Logic. Logic for Programming, Artificial Intelligence, and Reasoning, Mar 2012, Merida, Venezuela. pp.168-182. ⟨hal-00744100⟩