Improved K-means Algorithm And Its Application In The Cloud Task Allocation Strategy

Through in-depth study of the K-means algorithm, for the shortcomings of traditional K-means algorithm, the bacterial foraging algorithm and particle swarm optimization algorithm are applied to k-means algorithm, based on the MATLAB simulation environment, this paper proves that the improved K-means algorithm has significantly enhanced the clustering quality. Then through theoretical research on task scheduling algorithm based on K-means, we find that the scheduling capacity of the algorithm needs to be developed, so we use improved K-means algorithm for reasonable packet processing tasks, and then applying Min-Min algorithm for scheduling tasks, so that reaching or approaching the optimal scheme of task allocation. Through simulation experiment based on CloudSim cloud platform, the experimental results show that, the algorithm can not only reduce the overall time of task scheduling, but also improve the efficiency of task allocation and utilization ratio of system resources, the task completion time and load balance is better than Min-Min algorithm and Min-Min based on traditional K-means algorithm, research ideas will provide reference and help for the research of Cloud computing task allocation strategy in future.