E2E Delay Measurement For Campus's Network

Internet is an infrastructure of modern information society and also a complex giant system which plays an important role in our lives. The behavior of the Internet is changing constantly and unpredictable. With the development of network technology, many new network applications are emerging fast. Different applications have different traffic characteristics and QoS requirements. But the understanding of the network's behavior behind the development of the network and the traditional theoretical models can not adapt to the new network features. SNMP protocol can not meet the network monitoring and managing requirements. More and more QoS-sensitive business need to monitor the quality of service. Network technology provides effective reference data for the measuring problem. Based on the statistical analysis of the measured data we can make whole operation for the network. On the one hand, the assessment of current situation and c accurate description of network can be made effectively, thus determine what aspects of network needs to be improved and What questions need to be solved ,providing e reliable data for network optimization strategy. On the other hand that provides the basis for the future development of the feasibility of network planning.The thesis is mainly concerned with the measure ments for E2E network delay . First, the development of network measurement background and current development are introduced. Then it describes some of the basic concepts of network measurement in detail, for example, packet type, sampling and classification systems are introduced. Furthermore, on the campus network for the measurement of E2E delay, respectively, from the network topology, detection mechanisms and statistical inference techniques to make a detailed analysis of technology. In this paper, the cumulative generating function is used to deal with the measured E2E delay for access to all data sets the distribution of link delay. Through NS2 simulation, it proves the validity of the cumulative generating function. Finally this method, the campus network E2E actual delay measurement data analysis, and further speculated that the regional campus network bottleneck link for the campus network monitoring and optimization of providing strong support.