Resource scheduling becomes more complex as the introduction of virtualization technology in cloud computing. This paper proposed a resource scheduling model using the concept of resource service ratio as an object function and employing Estimation of Distribution Algorithms (EDAs) to solve this model. In scheduling model, the resource was abstracted to nodes with attributes, the scheduling result was evaluated by resource service ratio instead of the tasks completion time. In EDAs, two novel factors were introduced, one was the iteration times of best individual unchanged for reducing the iteration time, and another was share probability for improving the fitness. In experiment, when the tasks number is between 5 and 55 and the load rate is between 0.5 and 1.5, comparing with Max-min algorithm, static algorithm and random algorithm, the resource service ratio of EDA algorithm is improved on average by at least 1.004 and at most 1.793 times.
|Number of pages||12|
|Journal||International Journal of Digital Content Technology and its Applications|
|Publication status||Published - Mar 2012|
- Cloud computing
- Estimation of distribution algorithms
- Resource scheduling