Statistical multiplexing strategies for self-similar traffic
Linawati Linawati, Nyoman Putra Sastra · 2008
Traffic-shifting and shuffle methods are proposed as statistical multiplexing strategies for self-similar traffic sources. The purpose of these strategies is to decrease the burstiness of the traffic. This paper explores the Hurst parameter, and autocorrelation of the homogenous and heterogenous traffic. The simulation initially investigates the traffic with and without strategies implementation. Finally the comparisons between both methods are explored. The simulation results show that the strategies can decline the burstiness of the traffic. When contrasting the traffic performances achieved by both strategies, no large differences of H-parameters and autocorrelation coefficients of the traffic are observed.