Eliot: uma arquitetura para internet das coisas: explorando a elasticidade da computação em nuvem com alto desempenho

Abstract

The digital universe is growing at significant rates in recent years. One of the main responsible for this increase in the volume of data is the Internet of Things, which in a simplistic definition, is to uniquely identify objects electronically, track them and store their information for later use. To deal with such data load, solutions are needed at software, hardware and architecture levels. Studies conducted in this work show that the architecture adopted currently has limitations, specially regarding scalability. As scalability is a key feature to meet the growing demand for collection, processing and storage of data, this paper presents the architecture entitled Eliot, with proposals to precisely resolve the scalability and offers elasticity to the system. For this, is proposed the use of techniques of distributed databases, parallel processing and cloud computing, as well as restructuring of the current architecture. The results obtained after the implementation and evaluation of Eliot in a cloud computing environment demonstrate the feasibility, efficiency and reliability of this new proposed architecture. It was possible to improve performance by reducing response times and increased volume of requisitions processed and trafficked in the network, in addition to the reduction in connection failures and data communication.