A framework for mapping periodic real-time applications on multicomputers

A new strategy for multiprocessor scheduling of cyclic task graphs

“…An overwhelming majority of the work in the scheduling literature addresses directed acyclic task graphs (DAG), and only a few studies have been conducted in the scheduling of cyclic task graphs (Hanen and Munier, 1994;Reiter, 1968;Weinblatt, 1972) -also known as iterative computations. Most work on scheduling iterative computations has been conducted in the realms of instruction level parallelism (Knijnenburg et al, 2003) under the banner of software pipelining (Allan et al, 1995;Lam, 1988), and in real time systems (Shukla and Agrawal, 1994) (the retiming of the single dataflow graph paradigm (Calland et al, 1998;Chao and Sha, 1997;Jeng and Chen, 1992;Lucke and Parhi, 1993;Parhi and Messerschmitt, 1991;Piriyakumar et al, 1999;Tongsima et al, 2000;Wang and Hu, 1994;Wang and Parhi, 1995), but related problems also occur in other areas such as chemical process engineering and computer systems management (Sandnes, 2003).…”

Optimal dispatching and scheduling for both periodic and aperiodic jobs in cluster

“…We can construct the function of aperiodic workload admitted by the periodic job set J, denoted as S(t), by getting the minimal value of S 1 (t) and S 2 (t) at any time, shown in [4,6], [6,9] and [9,12], respectively. In Fig1.c, we can find that the real execution case of the aperiodic job accords with the arrangement of S(t), furthermore, the aperiodic workload executes first in each time slice.…”

“…Burchard [2] proposed admission control conditions of independent periodic jobs in a multiprocessor. Shukla and Agrawal [12] studied a policy to schedule periodic real-time applications with dependency on a multicomputer. Other periodic job scheduling algorithms on the multiprocessor have been also proposed in [5] [11].…”