2009 Volume 6 Issue 24 Pages 1737-1744
In this paper, a heuristic mapping algorithm which maps tasks, using priority lists and the crinkle moving pattern is proposed. To evaluate this algorithm, a set of real (i.e. Video Object Plan Decoder) and random applications have been used and the results have been compared. By reducing the number of hops between IP cores, the energy consumption and the completion time of the application (time which all tasks in the task graph execute wholly) have been optimized. Compared to other mapping algorithms, the algorithm execution time (due to its low complexity) is considerably lower.