Scheduling algorithms for application-level QoS support of continuous media streams

The Quality of Service (QoS) of real time applications is affected in a rat her direct manner by the timely delivery of Application-level Data Units (A DUs) which are typically mapped into several Network-level Data Units (NDUs ). As a consequence, it would be reasonable to associate a common deadline with all the NDUs in one ADU and develop a scheduling policy that aims to m aximise the number of ADUs that are delivered by their respective deadlines . In the first part of the paper, an ADU-based Earliest Due First (EDF) sch eduling policy is investigated under periodic ADU arrival processes which a re typical of continuous media streams. The proposed policy can be implemen ted in terms of a simple First ADU In - First ADU Out (FAI-FAO) policy and it is shown to improve on the ADU dropping rate induced by FIFO policies of comparable complexity. In the second part of the paper, the more complex a nd effective Size-Deadline Sensitive(SDS) scheduling policy is proposed and investigated under general ADU arrival processes. This policy is shown to be optimal in the sense that it maximises the number of complete ADUs sched uled for transmission at any point in time, by taking into consideration th e associated diverse deadlines and the ADU sizes and combining the principl es of two well known scheduling algorithms, the Earliest Due First (EDF) an d the Shortest Job First (SJF). It is argued and illustrated that the propo sed optimal scheduling leads in efficient actual service which under certai n arrival processes is shown to be the optimal. Numerical results are prese nted and the induced ADU and NDU loss rates under the proposed policy are c ompared against those induced by the standard EDF and SJF policies.