Rapporto tecnico, 2017, ENG

Design and optimisation of the flyfast front-end for attribute-based coordination. Preliminary version (revision 0.1)

Latella D.; Massink M.

CNR-ISTI, Pisa, Italy; CNR-ISTI, Pisa, Italy

Collective Adaptive Systems (CAS) consist of a large number of interacting objects. The design of such systems requires scalable analysis tools and methods, which have necessarily to rely on some form of approximation of the system's actual behaviour. Promising techniques are those based on mean-field approximation. The FlyFast model-checker uses an on-the-fly algorithm for bounded PCTL model-checking of selected individual(s) in the context of very large populations whose global behaviour is approximated using deterministic limit mean-field techniques. Recently, a front-end for FlyFast has been proposed which provides a modelling language, PiFF in the sequel, for the Predicate-based Interaction for FlyFast. In this paper we present details of PiFF design and an approach to state-space reduction based on probabilistic bisimulation for inhomogeneous DTMCs.

Keywords

Collective adaptive systems, Probabilistic model-checking, On-the-fly model-checking, Mean-field approximation, Discrete time Markov chains, Bisimilarity

CNR authors

Massink Mieke, Latella Diego

CNR institutes

ISTI – Istituto di scienza e tecnologie dell'informazione "Alessandro Faedo"