A model-driven engineering approach for the service integration of IoT systems
- PDF / 4,059,112 Bytes
- 18 Pages / 595.276 x 790.866 pts Page_size
- 21 Downloads / 180 Views
(0123456789().,-volV)(0123456789().,-volV)
A model-driven engineering approach for the service integration of IoT systems Darwin Alulema1,2 Rosa Ayala2
•
Javier Criado2
•
Luis Iribarne2
•
Antonio Jesu´s Ferna´ndez-Garcı´a3
•
Received: 11 September 2019 / Revised: 8 June 2020 / Accepted: 26 June 2020 Ó Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract With the development of IoT devices and web services, the objects of the real world are more interconnected, which allows applications to extend their characteristics in different fields, including industrial or home environments, among other possible examples such as health, trade, transport, or agriculture. However, this development highlights the challenge of interoperability, because devices are heterogeneous and use different communication protocols and different data formats. For this reason, we propose a model for point-to-point integration in three-layer IoT applications: (a) hardware, which corresponds to the physical objects (controller, sensor and actuator), (b) communication, which is the bridge that allows the exchange of data between a MQTT queue and REST web services, and (c) integration, which establishes a sequence of transactions to coordinate the components of the system. For this purpose, a metamodel, a graphic editor and a code generator have been developed that allow the developer to design IoT systems formed by heterogeneous components without having in-depth knowledge of every hardware and software platform. In order to validate our proposal, a smart home scenario has been developed, with a series of sensors and actuators that combined show a complex behavior. Keywords Model-driven engineering (MDE) Domain-specific language (DSL) Web services Integration pattern Internet of Things (IoT) Smart home
1 Introduction
& Darwin Alulema [email protected] Javier Criado [email protected] Luis Iribarne [email protected] Antonio Jesu´s Ferna´ndez-Garcı´a [email protected] Rosa Ayala [email protected] 1
Universidad de las Fuerzas Armadas ESPE, Sangolquı´, Ecuador
2
Applied Computing Group, University of Almerı´a, Almerı´a, Spain
3
Quercus Research Group, University of Extremadura, Ca´ceres, Spain
Nowadays, objects have greater processing, storage and communication capabilities, among others [9, 29], which has allowed them to be integrated with traditional enterprise level services [31, 34]. This development has a very important role in the Internet of Things (IoT), since it allows us to deal with the complexity of systems, increase the visibility of information and improve production performance [14, 35]. However, we face the problem of interoperability, because IoT devices can be heterogeneous, use different protocols (e.g., HTTP, MQTT, DDS and CoA) and manage different data formats (e.g., binary, XML, JSON and GIOP) [12, 23]. From the wide diversity of devices and platforms, two main challenges have emerged. Firstly, the heterogeneous nature of IoT information makes the task of interpreting that
Data Loading...
