A large fraction of traffic that the future B-ISDN will probably transport is made up of real-time services with stringent delay and delay jitter requirements. Since ATM networks do not provide time transparent links (i.e. constant delay links), a delay equalisation has to be provided in the adaptation layer or in user equipments. In this paper, we first present a source model analysis and then propose an analytical model that allows the evaluation of the end-to-end delay and of the relevant jitter; finally, we focus on the dimensioning of play-out buffers. The proposed model is validated with simulation results and we found a good agreement between analytical and simulation results. To make the study analytically tractable we use rather simple traffic source models and we make suitable simplifying assumptions. These assumptions are fairly general but the source models are still far from representing accurately certain kinds of real traffic. For this reason, we also carry out an additional simulation study (supported by some analytical arguments), by using real experimental MPEG and LAN traffic traces, to assess the system performance in real scenarios. (C) 1998 Elsevier Science B.V.
Dimensioning of play-out buffers for real-time services in a B-ISDN / N., Blefari Melazzi; Eramo, Vincenzo; Listanti, Marco. - In: COMPUTER COMMUNICATIONS. - ISSN 0140-3664. - STAMPA. - 21:11(1998), pp. 980-995. [10.1016/s0140-3664(98)00171-6]
Dimensioning of play-out buffers for real-time services in a B-ISDN
ERAMO, Vincenzo;LISTANTI, Marco
1998
Abstract
A large fraction of traffic that the future B-ISDN will probably transport is made up of real-time services with stringent delay and delay jitter requirements. Since ATM networks do not provide time transparent links (i.e. constant delay links), a delay equalisation has to be provided in the adaptation layer or in user equipments. In this paper, we first present a source model analysis and then propose an analytical model that allows the evaluation of the end-to-end delay and of the relevant jitter; finally, we focus on the dimensioning of play-out buffers. The proposed model is validated with simulation results and we found a good agreement between analytical and simulation results. To make the study analytically tractable we use rather simple traffic source models and we make suitable simplifying assumptions. These assumptions are fairly general but the source models are still far from representing accurately certain kinds of real traffic. For this reason, we also carry out an additional simulation study (supported by some analytical arguments), by using real experimental MPEG and LAN traffic traces, to assess the system performance in real scenarios. (C) 1998 Elsevier Science B.V.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.