01.1 - Graduação (Sede)
URI permanente desta comunidadehttps://arandu.ufrpe.br/handle/123456789/2
Navegar
2 resultados
Resultados da Pesquisa
Item Predição do consumo energético de dispositivos LoRa usando aprendizagem de máquina(2024-12-10) Pimentel, Henrique Pablo Pinheiro dos Santos; Araújo, Danilo Ricardo Barbosa de; http://lattes.cnpq.br/2708354422178489; http://lattes.cnpq.br/0078523045227122A Internet das Coisas (IoT) é um conceito em constante evolução que tem conquistado destaque tanto na comunidade acadêmica quanto na indústria. Dentro dela, o consumo energético é um fator fundamental para determinar o tempo de funcionamento dos dispositivos e a frequência necessária para realizar a manutenção deles. Este artigo investiga a aplicação de algoritmos de aprendizado de máquina para predição do consumo energético de dispositivos IoT-LoRa, permitindo estimar a duração da bateria dos dispositivos e sua autonomia. A metodologia considerou a criação de um conjunto de dados a partir de experimentos com placas de desenvolvimento Event stream processing (ESP32), capturando métricas como tempo de hibernação, tipo de conexão e consumo energético. Técnicas de Inteligência Artificial (IA) são então aplicadas para prever o consumo energético com base nessas variáveis. De acordo com os resultados obtidos, a melhor técnica para prever o consumo energético é a Decision Tree, com um coeficiente de determinação superior a 96%. O estudo contribui para processos decisórios que visam selecionar dispositivos IoT considerando a autonomia projetada para as baterias de tais dispositivos.Item Avaliação do consumo energético de smartphones em um sistema Smart Attendance(2019) Cabral, Jefferson de Carvalho; Nóbrega, Obionor de Oliveira; Lins, Fernando Antonio Aires; http://lattes.cnpq.br/2475965771605110; http://lattes.cnpq.br/8576087238071129; http://lattes.cnpq.br/1111059195459577Currently, the Internet of Things paradigm is being seen as the protagonist of the technological revolution in which the computer age lives. In this paradigm, the objects present in our daily lives will be connected to us and other objects. One of the major concerns in the (IoT - Internet of Things) universe is energy efficiency. Assuming that all objects can be connected to numerous networks simultaneously transmitting and receiving data, the wireless technologies must be able to act efficiently in the traffic of the information between the nodes without compromising the life of the batteries of those devices. In the educational context, these technologies aim at the developmentof intelligent classrooms as well as the creation of static applications that have thefunction of automating some manual activities carried out by teachers and students.One of these static applications is known as Smart Attendance, which aims to automate and solve the problem of the loss of 10% of class time for taking attendance by using technologies that promise a low power consumption. Although some solutions automate class attendance, the waste of 10% of class time still persists. This has commonly occurred due to some factors such as lack of visibility of the problem and inadequate use of the technology. Considering the last factor, this work aims to evaluate the energyconsumption of smartphones by taking roll in Smart Attendance systems using Bluetoothand Bluetooh Low Energy (BLE) technologies. Through the literature review, it was possible to analyze the work in the educational environment and to understand theIoT’s performance in this context, to understand the operation of the BLE protocol andto compare it with Bluetooth, to explore the work related to the automation of class attendance and to define a specification of the system data traffic, its message flowsand its operation. In order to evaluate the system, it was considered the simulation technique, which was chosen based on the main criteria used to select the systems performance evaluation techniques. Then, we used the OMNet ++ simulator with theINET framework to create a Mobile Ad Hoc Network (MANET) configured with Bluetooth and BLE characteristics, where the nodes of this network were the student and teacher smartphones. The results of this research point out that there are no quantitative differences between the Bluetooth and BLE technologies when compared within the metrics package error rate and signal-to-noise ratio. However, it proves the efficiency of the BLE over Bluetooth considering the energy consumption metric which makesBLE the most rational choice, among those evaluated, for the implementation of a Smart Attendance.