Skewness and Kurtosis in Function of Selection of Network Traffic Distribution

The available literature is not completely certain what type(s) of probabilitydistribution best models network traffic. Thus, for example, the uniform, Poisson,lognormal, Pareto and Rayleigh distributions were used in different applications.Statistical analysis presented in this paper aims to show how skewness and kurtosis ofnetwork traffic samples in a certain time interval may be criterions for selection ofappropriate distribution type. The creation of histogram and probability distribution ofnetwork traffic samples is also discussed and demonstrated on a real case.