Distribution and Allocation of Network Resources Based on Predictive Analyses of Time-Series Telecommunications Data

Abstract

With the fast development of different communication technologies, applications, and services, the adoption of advanced sensory and computing solutions, such as the various Internet of Things (IoT) and mobile computing solutions, is continuously growing. The massive adoption of mobile computing and IoT sensory devices encouraged the continuous growth of generated network traffic. Therefore, the selection of adequate solutions for efficient data processing became necessary. Despite numerous advantages arising from effective data processing, operators and enterprises working within the ICT domain have only limited amounts of available networking resources to store, process, and use valuable information extracted from large quantities of gathered data. In this paper, an optimal planning process and prediction of usage of network resources is examined. It takes into consideration the results of predictive modeling processes based on available sets of time series telecommunications data. The given forecasts enable effective selection of network architectures, as well as the distribution and allocation of network resources considering the cloud, edge, and fog networking concepts.