Poisson packet traffic generation based on empirical data

An algorithm for generating equivalent Poisson packet traffic based on empirical traffic data is presented in this paper. Two steps are required in order to produce equivalent Poisson packet traffic. Real traffic trace is analyzed in the first step. In the second step, a new equivalent synthetic Poisson traffic is generated in such a way that the first order statistical parameters remain unchanged. New packet inter-arrival time series are produced in a random manner using negative exponential probability distribution with a known mean. New packet size series are also produced in a random manner. However, due to specified minimum and maximum packet sizes, a truncated exponential probability distribution is applied.