Buffer Management and Priority Scheduling for ATM Traffic Shaping

Published

December 20, 1999

Author(s)

F Duran, T Lizambri, S A. Wakid

Abstract

The impact of buffer management and priority scheduling is examined in worst case scenarios when the aggregate incoming traffic is higher than the output link capacity of an ATM traffic shaper. To simultaneously reduce cell loss and extreme delay behavior for two or more classes of service, we show that an adaptive priority scheme is required. The priority of different service queues is dynamically modified by the adjustment of a variable Β, which allows the transition of a fixed priority scheme into one that takes into account the waiting time of cells already in buffer. Buffer memory management ensures that all service queues are guaranteed a minimum amount of memory, yet available memory can be shared between service queues if the load is unbalanced between them. This approach guarantees that no cells are lost under stress conditions until all buffer memory is exhausted. We provide insight into the general behavior of multiple classes of service competing for one output port. In addition, this work provides basic guidelines for the design of traffic shapers, and helps engineers understand the complexity of complying with various QoS requirements negotiated in a service contract.

Citation

International Workshop on Interactive Distributed Multimedia

Pub Type

Journals

Keywords

ATM networks, high-speed, QoS, quality of service, resource management

Citation

Duran, F. , Lizambri, T. and Wakid, S. (1999), Buffer Management and Priority Scheduling for ATM Traffic Shaping, International Workshop on Interactive Distributed Multimedia (Accessed October 31, 2024)