IoT-Networks group-based model that uses AI for workgroup allocation
[EN] This paper presents a centralized management architecture model for designing workgroup-based Internet of Things (IoT) and Internet of Everything (IoE) networks. The architecture establishes the organization of an object according to its functions and capacities in layers. From its model, it is...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2021 |
| País: | España |
| Institución: | Universitat Politècnica de València (UPV) |
| Repositorio: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Idioma: | inglés |
| OAI Identifier: | oai:riunet.upv.es:10251/181218 |
| Acceso en línea: | https://riunet.upv.es/handle/10251/181218 |
| Access Level: | acceso abierto |
| Palabra clave: | IoT IoT-smart architecture IoT-gateway M2M protocols ML classifiers Collaborative workgroups Network model Graph theory Smart IoT-networks INGENIERIA TELEMATICA |
| Sumario: | [EN] This paper presents a centralized management architecture model for designing workgroup-based Internet of Things (IoT) and Internet of Everything (IoE) networks. The architecture establishes the organization of an object according to its functions and capacities in layers. From its model, it is derived the design of the algorithms that give the network operation. These algorithms include the multi-protocol communication and interconnectivity algorithm, the routing algorithm, the resource sharing algorithm, and the grouping algorithm, all controlled by Artificial Intelligence (AI). The grouping algorithm consists of creating collaborative workgroups based on Machine Learning (ML) techniques that use the objects' features to allocating these within a workgroup that attends a type of service and within an architecture layer according to its capabilities. The model was tested with a simulation that shows the Machine-to-Machine (M2M) interaction between the devices involved in providing a service to a user within a Smart Home. This simulation uses an AI hosted within an IoT-Gateway to collect data on the features that define a connected object's functions and services. The extraction of the features is done using the Discovery of Functions and Services Protocol (DFSP) transported through an IoT-Protocol. With this information, the AI assigns a layer and a workgroup to a new object when it enters the network. The result of these tests can be used to know which ML technique has better accuracy. |
|---|