Statistical analysis of aggregated network traffic model from multiple on/off sources
Aleksei I. Plavan, Vyacheslav G. Kartashevsky, Irina Pozdnyak · 2023
In this paper, the aggregated random event sequence, representing network traffic, obtained by combining multiple random event streams with known statistical characteristics is analyzed. For this, an algorithm to generate random event streams representing a sequence of received network packets with a certain distribution law and correlation function is used. Random event streams simulating on/off traffic sources, where one part represents packet inter-arrival times and the other represents packet processing delays. Then two or more on/off source sequences are combined into one using original algorithm based on the Virtual Output Queueing (VOQ) approach used in network switches to mitigate head-of-line blocking issue, taking into account the priority of the input sequences. Then the statistical properties of the resulting aggregated sequence are analyzed.