Wasp Algorithm Used To Drive Dynamic Task Allocation In A Heterogeneous Computing System

In this paper, the Parallel Computing and the development of parallel computer is described, included the principle of parallel computing, system agencies, and classification. Heterogeneous computing systems have three-tier structure; included heterogeneous computer, communications and heterogeneous computing supported software. The heterogeneous computing parallel testing phase, parallel characteristics (type) disjunction stage and the task of mapping and scheduling phase are described. The heterogeneous computing system's tasks optimal scheduling problem is proposed, and introduced two heterogeneous computing task scheduling problems: algorithms static and dynamic algorithm. These two types of algorithm is common task scheduling and tasks will be to implement separation, for heterogeneous computing system which specific model is given or the task scheduling matrix.In this paper, the swarm intelligence is introduced into the scheduling of heterogeneous computing system, the swarm intelligence is described in detail about the definition of swarm algorithm and the three types of existing swarm algorithm, bees algorithm is a adaptive, distribution, flexibility and robustness algorithm.The swarm intelligence is introduced into the scheduling of heterogeneous computing system. A novel dynamic task-scheduling model for heterogeneous computing system is proposed, where the model describing the interaction between the wasp colony and the environment, with adaptability, flexibility and robustness.The paper describes the process as follows:First, a parallel computing and the related concept of parallel computer are introduced. Including the concept of parallel computing, parallel computing architecture, and the classification of parallel computing parallel computer category.Secondly, parallel computing and parallel computer on the basis of principle is described, heterogeneous computing and heterogeneous computing systems are described, including heterogeneous computing system's hierarchical structure, basic principles and classification. Analysis of a heterogeneous computing system in the model building needs to be resolved. And the key issues task scheduling, Task scheduling problems on the research status quo, of the task scheduling optimization algorithm that is based scheduling matrix, in accordance with the scheduling matrix assignment scheduling.Finally, the swarm intelligence is introduced into the scheduling of heterogeneous computing system. The swarm Intelligent Technology is introduced. The ant algorithm, the particle swarm algorithm and wasp algorithm. A novel dynamic task-scheduling model for heterogeneous computing system is proposed, where the model describing the interaction between the wasp colony and the environment. In this model, tasks according to the corresponding response values of the processors and the stimulating signals of the unfinished tasks.