A multiscale approach for a distributed event-based Internet of Things

Denis Conan (1, 2, 3, 4) , Léon Lim (2, 3, 4) , Chantal Taconet (1, 2, 3, 4) , Sophie Chabridon (1, 2, 3, 4) , Claire Lecocq (2, 3, 4)

1 ACMES-SAMOVAR - Architecture, Cloud continuum, formal Models, artificial intElligence and Services in distributed computing
2 TSP - INF - Département Informatique
3 SAMOVAR - Services répartis, Architectures, MOdélisation, Validation, Administration des Réseaux
4 CNRS - Centre National de la Recherche Scientifique

The Internet of Things paradigm calls for exchanging data among dynamic and heterogeneous producer and consumer entities at unprecedented scales. The approach used in this paper to address IoT heterogeneity and scalability is through the modelling of multiple heterogeneous scales along dimensions of the application domain. Then, the concepts of scale and dimension of multiscalability are mapped to the concepts of scope and graph of scopes of distributed event-based systems, which have long been recognised as enabling scalable and flexible communication in a space-, time- and synchronisation- decoupled way. This multiscale approach for a distributed event-based Internet of Things enables the modelling of the decentralisation with human-centred edge computing solutions placing control at the edges of the IoT by leveraging localised scalability. Our implementation and experimentations with the MuDEBS framework show that multiscoping helps to drastically diminish the number of exchanged messages for both subscriptions and notifications

Mots clés: Internet of Things; Multiscale distributed sys- tem; Distributed event-based system; Content-based routing; Localised scalability; Scoping; Multiscoping

Domaines: Informatique ubiquitaire

Denis Conan, Léon Lim, Chantal Taconet, Sophie Chabridon, Claire Lecocq. A multiscale approach for a distributed event-based Internet of Things. PICOM 2017 : 15th International Conference on Pervasive Intelligence and Computing, Nov 2017, Orlando, United States. pp.844 - 852, ⟨10.1109/DASC-PICom-DataCom-CyberSciTec.2017.142⟩. ⟨hal-01766251⟩