Showing papers by "Joost-Pieter Katoen published in 2007"

Bisimulation minimisation mostly speeds up probabilistic model checking

Abstract: This paper studies the effect of bisimulation minimisation in model checking of monolithic discrete-time and continuous-time Markov chains as well as variants thereof with rewards. Our results show that-as for traditional model checking-enormous state space reductions (up to logarithmic savings) may be obtained. In contrast to traditional model checking, in many cases, the verification time of the original Markov chain exceeds the quotienting time plus the verification time of the quotient. We consider probabilistic bisimulation as well as versions thereof that are tailored to the property to be checked.

Three-valued abstraction for continuous-time Markov chains

Abstract: This paper proposes a novel abstraction technique for continuous-time Markov chains (CTMCs). Our technique fits within the realm of three-valued abstraction methods that have been used successfully for traditional model checking. The key idea is to apply abstraction on uniform CTMCs that are readily obtained from general CTMCs, and to abstract transition probabilities by intervals. It is shown that this provides a conservative abstraction for both true and false for a three-valued semantics of the branching-time logic CSL (Continuous Stochastic Logic). Experiments on an infinite-state CTMC indicate the feasibility of our abstraction technique.

Counterexamples in probabilistic model checking

Abstract: This paper considers algorithms for counterexample generation for (bounded) probabilistic reachability properties in fully probabilistic systems. Finding the strongest evidence (i.e, the most probable path) violating a (bounded) until-formula is shown to be reducible to a single-source (hop-constrained) shortest path problem. Counterexamples of smallest size that aremostly deviating from the required probability bound can be computed by adopting (partially new hopconstrained) k shortest paths algorithms that dynamically determine k.

How fast and fat is your probabilistic model checker? an experimental performance comparison

Abstract: This paper studies the efficiency of several probabilistic model checkers by comparing verification times and peak memory usage for a set of standard case studies. The study considers the model checkers ETMCC, MRMC, PRISM (sparse and hybrid mode), YMER and VESTA, and focuses on fully probabilistic systems. Several of our experiments show significantly different run times and memory consumptions between the tools-up to various orders of magnitude--without, however, indicating a clearly dominating tool. For statistical model checking YMER clearly prevails whereas for the numerical tools MRMC and PRISM (sparse) are rather close.

Bisimulation and logical preservation for continuous-time markov decision processes

Abstract: This paper introduces strong bisimulation for continuoustime Markov decision processes (CTMDPs), a stochastic model which allows for a nondeterministic choice between exponential distributions, and shows that bisimulation preserves the validity of CSL. To that end, we interpret the semantics of CSL--a stochastic variant of CTL for continuous-time Markov chains--on CTMDPs and show its measuretheoretic soundness. The main challenge faced in this paper is the proof of logical preservation that is substantially based on measure theory.

Replaying play in and play out: synthesis of design models from scenarios by learning

Abstract: This paper is concerned with bridging the gap between requirements, provided as a set of scenarios, and conforming design models. The novel aspect of our approach is to exploit learning for the synthesis of design models. In particular, we present a procedure that infers a message-passing automaton (MPA) from a given set of positive and negative scenarios of the system's behavior provided as message sequence charts (MSCs). The paper investigates which classes of regular MSC languages and corresponding MPA can (not) be learned, and presents a dedicated tool based on the learning library LearnLib that supports our approach.

Verifying Concurrent List-Manipulating Programs by LTL Model Checking

Abstract: We present a novel approach to the verification of concurrent pointer– manipulating programs which operate on singly–linked lists. By abstracting from chains (i.e., non–interrupted sublists) in the heap, we obtain a finite–state representation of all possible executions of a given program. The combination of a simple pointer logic for expressing heap properties and of temporal operators then allows us to employ standard LTL model checking techniques. The usability of this approach is demonstrated by establishing correctness properties of a producer/consumer system and of a concurrent garbage collector.

Providing evidence of likely being on time: counterexample generation for CTMC model checking

Abstract: Probabilistic model checkers typically provide a list of individual state probabilities on the refutation of a temporal logic formula. For large state spaces, this information is far too detailed to act as useful diagnostic feedback. For quantitative (constrained) reachability problems, sets of paths that carry enough probability mass are more adequate. We recently have shown that in the context of discrete-time probabilistic processes, such sets of smallest size can be efficiently computed by (hop-constrained) k-shortest path algorithms. This paper considers the problem of generating counterexamples for continuous-time Markov chains. The key contribution is a set of approximate algorithms for computing small sets of paths that indicate the violation of time-bounded (constrained) reachability probabilities.

Motor: The MoDeST Tool Environment

Abstract: The MoDeST Tool Environment (MOTOR) is a tool to facilitate the transformation, analysis and validation of modest models. modest is a modelling language to describe stochastic real-time systems. MOTOR implements the formal semantics of modest and is designed to transform and abstract modest specifications such that analysis can be carried out by third-party tools. For the time being, a fragment of MoDeST can be model-checked using CADP. The main analytical workhorse behind motor is discrete-event simulation, which is provided by the Mobius performance evaluation environment. We are experimenting with prototypical connections to the real-time model checker Uppaal.

Abstraction of probabilistic systems

Abstract: Probabilistic model checking enjoys a rapid increase of interest from different communities. Software tools such as PRISM [13] (with about 4,000 downloads), MRMC [12], and LiQuor [2] support the verification of Markov chains or variants thereof that exhibit nondeterminism. They have been applied to case studies from areas such as randomised distributed algorithms, planning and AI, security, communication protocols, biological process modeling, and quantum computing. Probabilistic model checking engines have been integrated in existing tool chains for widely used formalisms such as stochastic Petri nets [6], Statemate [5], and the stochastic process algebra PEPA [11], and are used for a probabilistic extension of Promela [2].

motor :T hemodest Tool Environment

Abstract: The modest Tool Environment (motor) is a tool to fa- cilitate the transformation, analysis and validation of modest models. modest is a modelling language to describe stochastic real-time sys- tems. motor implements the formal semantics of modest and is de- signed to transform and abstract modest specifications such that anal- ysis can be carried out by third-party tools. For the time being, a frag- ment of modest can be model-checked using Cadp. The main analytical workhorse behind motor is discrete-event simulation, which is provided by the Mperformance evaluation environment. We are experiment- ing with prototypical connections to the real-time model checker Up- paal.