An Integrated Edge, Fog, and Cloud Computing Reference Architecture for Developing Data Ecosystems in Smart Cities

A smart city provides services to citizens through technological tools that analyze data from several subdomains for decision-making. Thus, a smart city platform is essential for integrating the city's subdomains and using data efficiently. To effectively tackle challenges in a multi-domain scenario, a flexible and adaptable solution that harnesses artificial intelligence efficiently is essential. This work presents UFCity, a data-centric, microservices-based solution with a three-layer network structure (edge, fog, and cloud computing). UFCity processes, stores, and analyzes data at each level, using an approach based on AI-equipped microservices on the cloud to extend services to citizens. A prototype was developed and tested in various usage scenarios to serve as a proof of concept. Furthermore, an experimental design $n=3^{k=3}$ was employed to evaluate the solution's performance based on message quantity, message group size for data fusion, and outlier removal methods for some elements. We conclude that UFCity is an advantageous alternative for managing city data, meeting requirements, and demonstrating high workload capacity.