Scheduling multiple independent hard-real-time jobs on a heterogeneous multiprocessor

Abstract

This paper proposes a scheduling strategy and an automatic scheduling flow that enable the simultaneous execution of multiple hard-real-time dataflow jobs. Each job has its own execution rate and starts and stops independently from other jobs, at instants unknown at compile-time, on a multiprocessor system-on-chip. We show how a combination of Time-Division Multiplex (TDM) and static-order scheduling can be modeled as additional nodes and edges on top of the dataflow representation of the job using SingleRate Dataflow semantics to enable tight worst-case temporal analysis. We also propose algorithms to find combined TDM/static order schedules for jobs that guarantee a requested minimum throughput and maximum latency, while minimizing the usage of processing resources. We illustrate the usage of these techniques for a combination of Wireless LAN and TD-SCDMA radio jobs running on a prototype Software-Defined Radio platform.

Hard-real-time scheduling of data-dependent tasks in embedded streaming applications

“…Moreira et al have investigated real-time scheduling of dataflow programs modeled as Homogeneous SDF (HSDF) graphs in [19,20,21]. They formulated a resource allocation heuristic [19] and a TDM scheduler combined with static allocation policy [20].…”

“…They formulated a resource allocation heuristic [19] and a TDM scheduler combined with static allocation policy [20]. Their TDM scheduler improves the one proposed in [3].…”

“…Our approach differs from [3,19,20] in: 1) using the periodic task model which allows applying a variety of proven hard-realtime scheduling algorithms for multiprocessors, 2) providing direct analysis of acyclic graphs for which the maximum cycle mean analysis can not be applied without converting the graph into a cyclic one, and 3) using the CSDF model which is more expressive than SDF graphs. Compared to [21], our approach differs in: 1) we derive an equation which always finds the actors periods that ensure periodic execution and scheduling.…”

“…Given µ and λ min computed in Example 3, we construct a task set ΓG = {(5, 8, 0), (2,8,8), (3,4,8), (2,6,20)}. We compute the minimum number of required processors to schedule ΓG according to Equations 5,6, and 7: M PAR = min x∈N {x : B is x-partition of Γ and Uy ≤ 1∀y ∈ B} = 3 ΓG is schedulable using an optimal scheduling algorithm on 2 processors, and is schedulable using PEDF on 3 processors.…”

“…Most of the existing run-time scheduling solutions assume applications modeled as task graphs and provide best-effort or soft-real-time services [22]. Few run-time scheduling solutions exist which support applications modeled using a MoC and provide hard-real-time services [10,3,19,20]. However, these solutions either use simple MoCs such as SDF/PGM graphs or use TimeDivision Multiplexing (TDM)/Round-Robin (RR) scheduling combined with heuristics.…”