Study On Network Traffic Analysis And Prediction In Secondary Vocational School

In secondary vocational schools the network traffic analysis is prerequisite condition for abnormal traffic monitoring,network traffic control and network performance evaluation.The network traffic prediction can be employed to predict the network traffic trend for schools.The predicted values can be the basis for the analysis of the behavior of users,the network application management and the traffic control.Firstly in our work,the characteristics of network flow is studied,then by analyzing the samples of the data acquisition from secondary vocational schools and studying the characteristics(self-similarity,long-range dependence(LRD),periodic,abrupt,chaos)of network traffic,we find that the network traffic in schools has the similar characteristics we described above.Secondly,we discuss the important problems existed in the process of predicting algorithm and constructing the predicted model by using the BP neural network structure,classification and network traffic prediction model.We mainly consider how to determine the number of nerve cells in input,hidden and output layers and how to select the activation functions and initial values etc.Finally,By calculating and analyzing,the number of nerve cells in input,hidden and output of BP neural network prediction model is determined.A proper activation function is selected and the samples of the data acquisition are normalized.By applying the training samples,the BP neural network prediction model of secondary vocational schools that we constructed is trained,and this prediction model is tested.our study shows that the BP neural network prediction model of secondary vocational schools in our paper can be applied in the prediction of network traffic in other secondary vocational schools.our model have solved the problem that we cannot predict the network traffic in daily network traffic management of secondary vocational schools.