|dc.description.abstract||The development of wireless networks and devices equipped with multiple sensors, and their connection to storage centres and data processing through the Internet, has led to the implementation of Internet of Things (IoT). The growing interest in the use of IoT technologies has generated the development of numerous and diverse applications. The proper functioning of these applications requires the control of enormous flows of data generated by mobile devices to the intelligent decision-making centres deployed in the cloud. Many of the services provided by the applications are based on knowledge of the location and profile of the end user. The contribution of this Doctoral Thesis focuses on the main components of the development of localization systems for IoT applications: (i) algorithms for processing wireless signals that enable indoor localization; and (ii) a distributed infrastructure to optimize the processing of these algorithms.
Regarding the first point, the work of this Doctoral Thesis focuses on the characterization of wireless signals behaviour. To this end, this point this research make use of Non-Linear, Linear and Ensembled models, and optimization using genetic algorithms in order to find the best transmission power levels setup for wireless transmitter. The results of this research show that the use of classification algorithms in the process of signal characterization is able to greatly improve the performance-based indoor localization mechanisms.
On the second point, this Doctoral Thesis proposes to make use of distributed systems, specifically, Fog Computing. This architecture has been designed to respond to the needs of a large number of applications. With the aim of improving its performance, this research proposes a robust and energy-efficient distributed infrastructure.||