Traffic flows ateb-prediction method with fluctuation modeling using Dirac functions

Abstract

In this paper a mathematical model for predicting traffic in computer networks based on heterogeneous differential equations describing the oscillatory motion was proposed. It is suggested that the solution to the equation should be modified on the basis of the sum of Dirac functions which have different arguments at different points of time to simulate bursts of network traffic. Based on this model, the soft-ware implementing network monitoring and forecasting of traffic on the network on the grounds of mathematical calculations and collected data was developed. Testing of models based on real network data from LPNU ACS Department was presented. Also, the simulation results have been compared with real traffic data. The results are illustrated in the form of both tables and plots.