Using mathematical forecasting methods to estimate the load on the computing power of the IoT network

Abstract

The size of the network, the number of nodes and connected devices are exponentially increasing due to the development of the Internet of Things (IoT). It becomes difficult to administer the monitoring of heterogeneous networks. It is necessary to use predictive models (Model Predictive Control) to deploy decision support systems related to the IoT network security. The article examines three popular mathematical forecasting methods, evaluates their accuracy and their using possibility in predictive models to solve the problem of assessing the load on the computing power of IoT devices, including servers and services.