Alocação de recursos baseado em prioridade de tarefas e consumo de recursos na computação em borda

Abstract

With low computational power in Internet of Things (IoT) devices, data must be processed and relevant information extracted in devices with higher processing capacity. Edge Computing emerged as a complementary solution to cloud computing, also known as Cloud-to-Thing continuum. Devices at the network edge have computational resources to handle the data processing and analysis that constrained IoT devices eventually cannot perform. Edge Computing allows data processing close to the end devices, and not only in the often far away Cloud, reducing latency for Internet of Things applications. However, the resource constraints of edge nodes, which have lower computational power than the cloud, make massive request processing challenging. The approach of this study evaluates the performance of different classification algorithms through the validation of metrics for classifying models. The most efficient classifier achieved an accuracy of 92% and amprecision of 90%. The results indicate good performance of using this classifier in the evaluated approach, with mean absolute error (MAE) and mean squared error (MSE) averages of 0.07 and 0.08, respectively. Additionally, using this selected classifier efficiently allocates resources for IoT requests. The results demonstrate that resource management occurs more efficiently with our proposed mechanism, with significantly lower resource utilization compared to the allocation method based on distance. Different scenarios were tested regarding the number of requests, edge nodes, and failure mechanism. Therefore, the proposed method can become a valuable tool for efficient resource management, with reduced computational cost and efficient resource allocation.