An efficient ATM voice service with flexible jitter and delayguarantees

摘要

A transmission and multiplexing strategy appropriate for voicenover asynchronous transfer mode (ATM), called delayed frame queueingn(DFQ), is proposed. This frame-based strategy has features in commonnwith the synchronous transfer mode and is thus well suited to servicensynchronous applications such as voice, while retaining the statisticalnmultiplexing capabilities of ATM. In particular, the DFQ servicendiscipline can provide explicit and nontrivial bounds for queue delaynand jitter, for both bursty as well as continuous traffic streams.nFurthermore, the DFQ discipline can combine a wide range of delay andnjitter bounds while also managing the distribution of quality of servicenviolations among the traffic streams when congestion occurs. Jitterncontrol is performed at the network periphery and thus does notnnegatively influence multiplexing gain at intermediate nodes. Thisnefficient strategy has major implications in terms of the preferrednalternatives chosen by clients when implementing source clock recoverynfor voice. DFQ allows the entire range of implementation alternativesnfor voice over ATM to be appropriately serviced, such as ATM adaptationnlayer types 1 and 2 (AAL1/2), adaptive playout, and immediate playout