H. Beitollahi and S.G. Miremadi (Iran)
Partitioning scheduling algorithms, real-time systems,multiprocessor systems, fault-tolerance.
This paper presents the performance analysis of several best-known partitioning scheduling algorithms in real-time and fault-tolerant multiprocessor systems. To do this, multiple versions of tasks are executed on different processors. Both static and dynamic scheduling algorithms are analyzed. In the case of static scheduling algorithms, rate-monotonic (RM) scheduling policy is considered. In the dynamic scheduling algorithms, the scheduling policies are rate-monotonic and earliest-deadline-first (EDF). Partitioning scheduling algorithms which are studied here are heuristic algorithms that are formed by combining any of the bin-packing algorithms with any of the schedulability conditions for the RM and EDF policies. A tool is designed and implemented for performance evaluation. The simulation results show that among several evaluated algorithms, the rate-monotonic first-fit decreasing utilization (RM-FFDU) is the best static and the earliest-deadline-first best-fit (EDF-BF) is the best dynamic partitioning scheduling algorithms in real-time and fault-tolerant multiprocessor systems.
Important Links:
Go Back
