Dynamic Window-Constrained Scheduling of Real-Time Streams in Media Servers
IEEE Transactions on Computers
Firm Real-Time System Scheduling Based on a Novel QoS Constraint
IEEE Transactions on Computers
Event-driven scheduling for dynamic workload scaling in uniprocessor embedded systems
Proceedings of the 2006 ACM symposium on Applied computing
DLB: a novel real-time QoS control mechanism for multimedia transmission
International Journal of High Performance Computing and Networking
Proceedings of the 2011 International Conference on Communication, Computing & Security
Energy minimization for real-time systems with (m; k)-guarantee
IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Hi-index | 0.01 |
Window-Constrained Scheduling is one of the task modelsproposed in recent years for scheduling periodic real-timetasks where service must be guaranteed in only a fractionof the periods.In RTSS'2000, a Dynamic Window-ConstrainedScheduling (DWCS) algorithm was proposedby West and Poellabauer [17] for multiplexing multiplepacket streams where the number of consecutive packetlost must be bounded.In this paper, we show that theDWCS algorithm can fail for arbitrarily low aggregate utilizationrates of the packet streams.We shall show thatWindow-Constrained Scheduling is NP-hard in the strongsense.However if the execution time of all jobs is of unitsize, as might be modelled in some packet stream applications,then a schedule must exist as long as the aggregateutilization rate is no more than the number of processors(packet switching resources).Some subclasses of theWindow-Constrained Scheduling for unit-size jobs can betransformed to the well known Liu and Layland model orthe more recent Pfair model and can therefore be scheduledoptimally and at low scheduling cost.In consideringthe general unit-size-job case, we note that non-proportionateprogress in scheduling Window-Constrainedtasks is often the cause of unschedulability, no matter howlow the aggregate utilization rate is.We shall define thenotion of Pfairness in relation to the Window-ConstrainedScheduling model so as to quantitatively describe the conceptof strictly proportionate progress.An EDF (Earliest-Deadline-First) based algorithm will be defined for theWindow-Constrained Scheduler problem. This algorithmis computationally efficient, on-line, and Pfair.A sufficientschedulability test for this EDF-based algorithm is proved.