Graph Metrics for Predicting Speedup in Static Multiprocessor Scheduling
This paper presents a set of metrics for estimating the speedup achievable in static multiprocessor scheduling using a previously introduced Genetic Algorithm (GA) approach. This is of major importance because, although conventional wisdom suggests that metaheuristics such as GAs have the potential to improve over standard heuristics, little research has been conducted on characterizing the sorts of graphs that they should excel at. We describe several metrics and illustrate that four of them can predict the speed up with an accuracy of almost 90%.
KeywordsGenetic Algorithms Scheduling Graph Partitioning
Unable to display preview. Download preview PDF.
- 1.Khan, A.A., McCreary, C.L., Jones, M.S.: A comparison of Multiprocessor Scheduling Heuristics. In: International Conference on Parallel Processing, vol. 2, pp. 243–250 (1994)Google Scholar
- 2.Wu, A.S., Yu, H., Jin, S., Lin, K., Schiavone, G.: An Incremental Genetic Algorithm Approach to Multiprocessor Scheduling. IEEE Transactions on Parallel and Distributed Systems 15(9) (2004)Google Scholar
- 6.Yang, T., Gerasoulis, A.: A Fast Scheduling Algorithm for DAGs on an Unbounded Number of Processors. In: 5th ACM International Conference on Supercomputing, pp. 633–642. Association of Computing Machinery, New York (1991)Google Scholar
- 7.Sheahan, A., Ryan, C.: A Transformation-Based Approach to Static Multiprocessor Scheduling. In: Gecco 2008: Proceedings of the 10th Annual Conference on Genetic and Evolutionary Computation, pp. 1041–1048. Association of Computing Machinery, New York (2008)Google Scholar