Skip to Main content Skip to Navigation
Conference papers

Safe Composition in Middleware for the Internet of Things

Abstract : The Internet of Things (IoT) connects sensors, actuators and autonomous objects interacting with each other. These devices are represented by web services. Web services composition often involves conflicts between systems having access to shared devices. In our component-based middleware, our solution allows managing access to shared devices, by generating specific constraint components which guarantee the respect of some predefined composition and adaptation constraints. IoT environments are dynamic; our solution ensures adaptation to its changes by using new generated constraint components and inhibitors to deal with the appearance and disappearance of devices/applications. The main contribution in this work is the definition of a new language DCL (Description Constraint Language) that helps to generate our constraint components by describing generic constraints that must be verified on accesses to shared devices. The whole approach and its associated tools rely on the synchronous paradigm, since it has a well-established formal foundation allowing automatic proofs, and interface with most model-checkers. We can then prove and guarantee a safe composition at runtime for our IoT applications.
Complete list of metadata

Cited literature [12 references]  Display  Hide  Download
Contributor : Annie Ressouche Connect in order to contact the contributor
Submitted on : Wednesday, December 2, 2015 - 2:42:25 PM
Last modification on : Thursday, January 20, 2022 - 4:16:14 PM
Long-term archiving on: : Saturday, April 29, 2017 - 2:51:35 AM


Files produced by the author(s)




Ines Sarray, Annie Ressouche, Daniel Gaffé, Jean-Yves Tigli, Stéphane Lavirotte. Safe Composition in Middleware for the Internet of Things. Middleware for Context-aware Applications for Internet of thing (M4IoT), Dec 2015, Vancouver, Canada. ⟨10.1145/2836127.2836131⟩. ⟨hal-01236976⟩



Les métriques sont temporairement indisponibles