Avaliação de desempenho de computação em nuvem e névoa: estudo de caso de um sistema de automação residencial com redes neurais artificiais

Abstract

With the increase in the number of IoT devices, some of the disadvantages of implementing central management using cloud computing are beginning to appear. Issues such as high cost to store data and run applications, low privacy in data transmission and high latency can make it difficult to use, leading to a bad experience for users. Considering these characteristics of cloud computing, the concept of fog computing emerged. Fog computing is located closer to the edge of the network to reduce latency and make room for applications and services to take advantage of geographically closer processing. This work explores, in an applied way, in the context of smart homes, the impact of latency between fog computing and cloud computing using an application of artificial neural networks to predict the behavior of a person that lives in the house. The results showed that fog computing can serve as a viable alternative to cloud computing in current home automation deployments, thus reducing implementation costs.