Campus Network Behavior Analysis And Traffic Prediction

As the scale of the campus network and diversified development of network applications,the network traffic perdition had to be one of the main problems under complex,diverse network environment.Due to the increasing size of the campus network users,the difficulty of management of campus network increasing day by day.At the same time,some of problems band been disclosed.In this case,to understand the behavior of the campus network users is particularly important for the campus network management.Therefore it is needed to do the campus network flow characteristics analysis and prediction research.As the network bandwidth cannot meet the growing business needs of the school,the network service cannot be guaranteed.There are delay,lack of bandwidth and other issues.The rapid development of network communication technology demand high quality of network service,traffic control and network management.The analysis and forecast of the flow is the basis of network management and performance analysisIn this paper,we take campus network of Guangdong College of Pharmacy for example,analyzed several main characteristics of the network traffic and its impact on the performance of network traffic and network operations.In this case,the campus network requires timely and effective understanding of network traffic characteristics and the Internet in schools implementing flow measurement and control strategies implemented to ease network congestion to optimize the school network resources.Consideration of the current issue of network management school,it require measurements of the flow characteristics of the period on the basis of the user,sub-regional development and implementation of various traffic control strategies,and acquire the campus network user behavior law by vast amounts of log data.Finally,according to the results of the statistical analysis in MATLAB platform test gives examples of network traffic forecast method-BP neural network prediction algorithm.Explore the behavior of Internet users in a specific school of law by BP neural network model,which can achieve standardized management,load balancing,and enhance information security purposes.(1)In this paper,based on data characteristics campus network users,to loss of information within the allowable range as a precondition,improve the efficiency of data filtering and analyzing network prediction involving relevant content.(2)Through the campus network users to use network services,data access connection acts as a variable,distribution,model laws,such as the frequency of training within the time limit and so on,network traffic logs on a statistical analysis of the data,statistical data and network traffic data to understand the basic characteristics of the distribution of time and then get the traffic data rule by the analysis of user network traffic.(3)By the summary of advances in network traffic prediction algorithm,and by the currently exists BP neural network prediction algorithm model to get the regularity of network traffic analysis and forecasting.Finally,to get the BP neural network algorithm feasibility of network traffic forecasting results in theory.