A New Metric for On-Line Scheduling and Placement in Reconfigurable Computing Systems

Abstract

AbstractReconfigurable computing systems use reconfigurable processing unit in conjunction with a processor to make us able to execute tasks in a true multitasking manner. This leads to highly dynamic allocation situations. To manage such systems at runtime, a reconfigurable operating system is needed. On-line scheduling and placement algorithms are the main parts of this operating system. In this paper, we present a technique for on-line integrated scheduling and placement which focuses on on-line, real-time and non-preemptive reconfigurable computing systems. The main characteristic of our method includes using a new metric for selecting the best feasible placements for arriving tasks. This new metric is based on temporal and spatial constraints. A large variety of experiments has been conducted on the proposed algorithm using synthetic and real tasks. Obtained results show the benefits of this algorithm.KeywordsSchedule AlgorithmOnline SchedulePlacement AlgorithmFuture TaskReal TaskThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.