Per-node delay assignment strategies for real-time high speed networks

A variety of packet scheduling strategies have been proposed to guarantee the application's end-to-end delay requirements in real-time networks. The schemes, however, do not address specifically the mechanisms used to assign a per-node delay or service rate in order to meet the required end-to-end delay. In this paper, using a methodology for computing feasible per-node delays along a routing path, we describe different strategies to assign a delay value for each node so that the end-to-end delay requirement is satisfied. The performance of these schemes for different network environments is discussed.