S. Pawaskar and H.H. Ali (USA)
Scheduling, Energy Awareness, Heuristics, Parallel Processing, Optimal algorithms.
Scheduling dependent tasks is one of the most challenging problems in parallel and distributed systems. It is known to be computationally intractable in its general form as well as several restricted cases. An interesting application of scheduling is in the area of energy awareness for mobile battery operated devices where minimizing the energy utilized is the most important scheduling policy consideration. A number of heuristics have been developed for this consideration. In this paper, we study the scheduling problem for a particular battery model. In the proposed work, we show how to enhance a well know approach of accounting for the slack generated at runtime due to the difference between WCET (Worst Case Execution Time) and AET (Actual Execution Time). Our solution exploits the knowledge gained about the AET of the tasks after the first period, to come up with EET (Expected Execution Time). We then use the EET as an input for the next period to use as much slack as possible and to eliminate wastage of slack generated. This happens because WCET is used to determine if a task should be executed at runtime. Dynamically adjusting the run-queue to use EET as a feedback, which is based on the previous period’s AET eliminates wastage of the slack generated. Based on the outcome of the conducted experiments, the proposed algorithm outperformed or matched the performance of the 2-Phase dynamic task scheduling algorithm and the run-queue peek algorithm all the time.
Important Links:
Go Back
