Research And Application Of Campus Network Traffic Monitoring System

As the computer network gets expanding and more pervasive, it becomes more important to our life. The college campus network is the forefront of computer network construction, Zhaoqing University campus network has achieved a "gigabit fiber backbone, 100M exchange desktop", the three-tier structure of the network-building objectives, the network continues to expand the scale, network traffic is becoming increasingly complex, the issue of network management increasingly heavy. Development of an effective network traffic monitoring system, the campus network is the network planning and optimization of the foundation, to improve network management capability is of great significance.First, this essay analyses different types of network traffic on today's campuses. It introduces and compares different methods and technologies in measuring traffic and points out the necessity of measuring network traffic. Second, the essay discusses the SNMP protocol as well as related technologies and standards used in measuring traffic. Third, the traditional network traffic management system operation model were discussed, The idea it is really practical and meaningful to use browsers to administrate the network traffic. This will enable us administrate the network anywhere at any time. Fourth, the network flow for the method of study that network traffic flow must consider the time sequence, and ARIMA (Auto Regressive Integrated Moving Average) time series models to reflect this characteristic.This paper based on the network traffic management of the development trend of using layered architecture thinking, Design and Implementation of a Web-based SNMP management of network traffic monitoring systems.It uses SNMP protocol to collect MIB data, and then build a network traffic database. All high-level applications of this system is build on the analysis to database. This paper presents a network traffic prediction model which bases on ARIMA(p,d,q). This model solves the existing problems of random disturbance and too much uncertainty in network traffic prediction. It achieves better prediction results. With the prediction of network traffic and bandwidth utilization at any given time, this paper presents a more valuable equipment performance alarming methods. This model gets satisfactory results in practical uses, and improves the efficiency of network administration. It makes the network administration more flexible and scalable. At the end, we implement network traffic monitoring on the major and typical network equipments in Campus Network of Zhaoqing University, and made detailed analysis on gathered data.