A Network Traffic Network Prediction Model with K-Means Optimization Algorithm

Abstract

Data storage and computation on cloud servers can handle many gigabytes of data, but their network traffic will also be heavy. Researchers have developed several models for predicting network traffic in order to reduce the communication pressure on cloud servers. However, the existing models are not accurate enough to be applied to the cloud server. To deal with this problem, a network traffic prediction model (NTPM) with the K-means optimization algorithm is presented in this paper, which clusters network traffic data from cloud servers by the K-means optimization algorithm, and then SVM is used to train the model. Our study shows that compared with a recent NTPM that predicted network traffic accurately, the proposed model provides better network traffic prediction and is better suited to cloud servers.