This work is supported in part by the National Science Foundation through grants
NCR-9624106 (CAREER), ANI-9730103, and ANI-9903001.
Abstract
A novel framework, called JoBS (Joint Buffer Management and Scheduling), is presented for reasoning about relative and absolute per-class service differentiation in a packet network without information on traffic arrivals. JoBS has two unique capabilities: (1) JoBS makes scheduling and buffer management decisions in a single step, and (2) JoBS supports both relative and absolute QoS requirements of classes. JoBS is presented in terms of the solution to an optimization problem. Numerical simulation examples, including results for a heuristic approximation of JoBS, are presented to illustrate the effectiveness of the approach and to compare JoBS to existing methods for loss and delay differentiation.