Prediction-Based Elastic Load Balancing Mechanism in Cloud Environment

An elastic load balancing mechanism in cloud computing environment is studied in this paper.The mechanism that uses kNN (k-Nearest Neighbors) and Naive Bayes classification algorithms in machine learning can effectively solve the problem of resource allocation lag by predicting the future load trend on the basis of analysis and study of historical data.And taking into account the cross regional nature of the cloud computing environment, applications will be deployed to the computing nodes closer to the user to reduce user access time.Finally, we verify the feasibility and effectiveness of the proposed mechanism through some experiments.