Abstract: Aiming at the problem of low efficiency and high cost in cloud task scheduling, a new cloud task scheduling algorithm based on modified K-means clustering is proposed. Based on the hardware properties of the virtual resources, the improved clustering algorithm is used to cluster the virtual resources. The task preferences are computed, so that different preferences of the task could select resources in different clusters. Taking into account the scheduling cost problem, the task scheduling is performed on each cluster by using the improved Min-min algorithm. We improve the K-means clustering algorithm aiming at the problem that the initial cluster centers are randomly selected and it is easy to fall into local optimal solution. Finally, the cloud simulation platform CloudSim is used to carry out the experiment, and the results show that the proposed algorithm can improve the efficiency compared with the non-clustering scheduling algorithms.

Key words: cloud computing, K-means clustering, scheduling, CloudSim