A cloud task scheduling algorithm based on QK-mean clustering

Guozhu Zhao, Liang Ma, Ruibin Zhao

COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12A) 221-227

School of Computing & Information, Chuzhou University, Chuzhou Anhui 239012, China

When classifying resources incloud computing environment using the idea of clustering, the information entropy of resources’ attribute can reflect the degree of significance in clustering process. Using information entropy, a task scheduling algorithm based on QK-mean clustering is proposed. We calculate the degree of significance of cloud resources’ attributes, then apply K-mean clustering algorithmto classify the cloud resources according to the degree of significance of attributes, and we create Resources K-tree to store the process and result of the clustering. In this way, we transform the task scheduling process into the process of searching a suitable leaf node in Resources K-tree. The experimental results show that the QK-mean scheduling algorithm can effectively improve the efficiency of cloud task scheduling