BUFFER LOSSES VS DEADLINE VIOLATIONS FOR ABR TRAFFIC IN AN ATM SWITCH- A COMPUTATIONAL APPROACH

The B-ISDN will carry a variety of traffic types: the Variable Bit Rat e traffic (VER), of which compressed video is an example, Continuous B it Kate traffic (CBR:), of which telemetry is an example, Data traffic , and Available Bit Rate traffic (ABR) that represents aggregate data traffic with very limited guarantees on quality. Of these, VER and CBR have timing constraints and need synchronous bandwidth; data traffic is relatively delay insensitive. In this paper, we consider the VER, D ata and ABR traffic types and obtain the cumulative distribution funct ion (cdf) of the queueing delay experienced by a burst of ABR traffic in the output buffer of an ATM switch. The cdf is used to trade off bu ffer loss probabilities against deadline violation probabilities throu gh adjusting the buffer size and (delay) deadline values. Large buffer s result in low losses but queueing delays can become excessive and ca use a high level of deadline violations. Both losses and violations ar e detrimental and an operating point must be chosen to achieve a balan ce. In this paper we study the nature of the trade off. We develop a s tochastic Petri net model assuming periodic burst arrivals for VER and Poisson arrival processes for the Data and ABR traffic types at the b urst level; and solve the model analytically (numerically) using a dec omposition approach. This decomposition, along with the inherent decom posability of the tagged customer approach for obtaining the cdf opens up a possibility of carrying out fast computations using a parallel m achine for selecting the operating point each time that a call is admi tted.