Optimal Task Partitioning Model in Distributed Heterogeneous Parallel Computing Environment

摘要

Parallel computing systems compose task partitioning strategies in a true multiprocessing manner. Suchsystems share the algorithm and processing unit as computing resources which leads to highly interprocess communications capabilities.We focus on real-time and non preemptive systems. A large variety ofexperiments have been conducted on the proposed algorithm. Goal of computation model is to provide arealistic representation of the costs of programming.Thepaper representsthe optimal iterative task partitioning scheduling in the distributed heterogeneousenvironment. Main goal of the algorithm is to improve the performance of theschedule in the form ofiterationusing results from previous iterations. Thealgorithm first usesthe b-level computation tocalculate the initial schedule and then improve it iteratively.The results show the benefit of the taskpartitioning. The main characteristics of our method are optimal scheduling and strong link betweenpartitioning, scheduling and communication. Some important models for task partitioning are alsodiscussed in the paper. We target the algorithm for task partitioning which improvethe inter processcommunication between the tasks and use the recourses of the system in the efficient manner. The proposedalgorithm contributes the inter-process communication cost minimization amongst the executing processes.This paper is the extendedversion of [15].