Performance evaluation of software defined networking based on openflow protocol

This paper presents the results of traffic analysis obtained in the processing of streams in SDN. It is found that there is no significant effect on the statistical characteristics of stable mode processing in SDN. In the congestion mode, significant correlations arise in the sequence of time intervals between packets. Analytical expressions for the estimation of average values of packet delay time, delay variation, and queue length based on the hyperexponential distribution approach are obtained, and the H2/D/1 system for SDN is used. A methodology for estimating traffic service quality parameters in an SDN network is developed.