Spline-Extrapolation of Video Traffic of IoT-Devices Based on Various Cubic Splines

Abstract

The solution to the problem of predicting the characteristics of real-time video traffic of IoT network objects is considered. The classification of traffic of IoT network objects depending on the type of IoT/5G network services is proposed. For the considered video traffic in real time, the spline-extrapolation method based on the cubic Hermite spline was used. Comparison of the results of predicting the characteristics of video traffic using the spline-extrapolation method based on various cubic spline functions (cubic spline, cubic B-spline and cubic Hermite spline) is carried out. It is established that the obtained results of short-term forecasting of video traffic will provide necessary forecast accuracy and the possibility of using it for various IoT applications in order to prevent network congestion, especially in conditions of maximum network load.