Open Archive Toulouse Archive OuverteOATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible To cite this version: Farah, Zoubeyr and Ait Ameur, Yamine and Ouederni, Meriem and Tari, Kamel A correct-by-construction model for asynchronously communicating systems.Abstract The design and verification of distributed software systems is often hindered by their ever-increasing com-plexity and their asynchronous operational semantics. This article considers choreography specifications for distributed systems to reduce that complexity. We use labelled state-transitions systems as ground model for both choreographies and the corresponding distributed systems. Based on Event-B method, we propose a stepwise correctby-construction model to build asynchronous distributed systems which a pri-ori realise their choreographies. We rely on a sufficient and necessary realisability condition and we apply several refine-ment steps w.r.t. that condition to generate the distributed peers. The first refinement returns peer behaviours obtained by synchronous projection. The previously computed system is then refined into its asynchronous version using unbounded FIFO buffers. We prove, thanks to invariant preservation, that a sequence of exchanged messages is preserved at each refinement step. We provide a formalised proof of a realis-ability algorithm for deterministic choreographies. Besides that, our contribution is twofold: the approach is a priori and the problackposed solution scales up to any number of peers communicating with each other.
Communicating Finite State Machines (CFSM) lack the high level syntactic and structural abstractions of Communicating Complex State Machines (CCSM), such as nesting and encapsulation, to model highly complex protocols that are likely to arise in web services environments. The incorporation of these features in a protocol specification model would require the design of a new validation technique to efficiently check for protocol errors, such as deadlocks and non-reachable transitions. A reachability graph is used to represent the execution states of the protocol and to verify their consistency. In this paper, we propose a new validation technique for protocols modeled with complex FSM, called RLRA (Reverse Leaping Reachability Analysis), which enables the detection of all deadlock errors. It is a backtracking approach, which first identifies an initial set of suspected states, those possibly containing deadlocks, then refines this set to those likely to cause deadlock, and finally backtracks through the graph while checking for errors until the root state of the protocol is reached. Leap graphs are employed to prune the number of execution states examined, and thereby mitigate the combinatorial explosion of the state space. Extensive tests and comparisons were performed, which show the effectiveness of our technique.
Controlling access to data is one of the primary purposes of security, especially when it comes to dealing with safety critical systems. In such systems, it is of paramount importance to rigorously define access control models. In this article, a correct-by-construction specification of RBAC using the Event-B formal method is proposed. The specification defines closely the model properties with the behavior aspect of RBAC as guards of events, which allows applying a priori verifications. Accordingly, the resulted specification is correct-by-construction and avoids the combinatorial explosion problem. As well, a number of refinement operations are performed leading to a specification with several abstraction levels, where each level implements selected RBAC entities. The approach is illustrated by an instantiation of a healthcare system.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.