Gerenciamento proativo baseado na análise computacional dos Nodos Fog que integram Smart Classroom

• Main Author: Fischer, Ivania Aline
• Publication Date: 2020
• Format: Master thesis
• Language: por
• Source: Manancial - Repositório Digital da UFSM
• dARK ID: ark:/26339/0013000011x9k
• Download full: http://repositorio.ufsm.br/handle/1/22278
• topic: Gerenciamento da computação em névoa; Computação em névoa; Sala de aula inteligente; Análise computacional nodos da computação em névoa; Redundância de nodos da computação em névoa; MagProFog; Management fog computing; Fog computing; Smart classroom; Computational analysis nodes fog; Fog redundacy; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

The Internet of Things (IoT) has presented the development of applications focused on smart cities, transport, health and education with Smart Classrooms. That said, the increasing rates and high demands for real-time services, and as a consequence, the latency of communications has become an important issue. As Cloud Computing (CC) does not provide the necessary support for this type of service, Computing in Fog (FC) is highlighted as a means of helping the latency problem faced. In addition, Fog subsidizes the use of basic characteristics of IoT environments such as mobility treatment, distributed computing layout and device heterogeneity. In this work, the focus is related to IoT in education specifically in the integration of a Smart Classroom with Fog in order to acquire greater agility in communications and decrease the use of bandwidth, which is a precarious factor in public education institutions. With a focus on Fog and its proper functioning in Smart Classroom, this work has the objective of acquiring satisfactory levels of operationality. To obtain this item a model of proactive management of Fog nodes is proposed, called MagProFog which is divided into three modules, the Database module, Monitoring module (developed in the FogTorch simulator) and the Management Interface module. The management has the purpose of identifying and preventing possible future irregularities, its identification is acquired by Fog is overhead rate, correlated with the storage and processing usage rates. The prevention of problems caused by Fog is irregularity is intrinsically correlated with obtaining possible neighboring Fogs nodes for replacement before the irregularity impairs the performance of the Smart Classroom. The definition of the Fog node that best meets the objective above is obtained through a computational analysis of performance, the use load of Fog and the latency of the replacement process. The work validation tests are applied in a simulation environment (using the three management modules), where the communication data used by Smart Classroom is real acquired by its prototyping. In this way, we seek to identify the operational levels of Smart Classroom acquired in the application of a set of Fogs irregularities resolution policies coupled with the proposed computational management and analysis. The acquired results are promising, when applied to an environment that demonstrates irregularity, a percentage of 100% of operability is achieved in a set of Smart Classrooms on a Smart Campus.