Characterization of Specific Applications using Statistical Methods in Distributed Systems

The problem of mapping tasks and communications onto multiple machines and networks in a heterogeneous computing environment has been shown to be NP complete. Therefore, the development of heuristic techniques to find near-optimal solutions is required. Many different types of mapping heuristics have been developed in recent years. However, selecting the best heuristic to use in any given scenario remains a difficult problem. Moreover, it is not possible to make a general mapping in a heterogeneous computing environment. In this paper we propose to characterize classes of applications with the objective of predicting their behavior. Using the insight provided by the characterization, we achieve a more realistic mapping for specific applications. Key-words: Heterogeneous system, scheduling, heuristics, dynamic mapping, Distributed Simulation