Finding hierarchical heavy hitters in network measurement system
Proceedings of the 2007 ACM symposium on Applied computing
Hi-index | 0.00 |
Traffic measurement and analysis, as one of the important methods of understanding and characterizing network, can provide significant support for network management. After a brief introduction of a novel NP (Network Processor)-based architecture of traffic measurement, the paper presents the detailed analysis results of the traffic collected from the Gigabit link connecting Tsinghua University campus network (in short, TUNET) to its upstream ISP, China Education and Research NETwork (in short, CERNET). Then, the paper comprehensively analyzes the traffic from multi-dimension viewpoints, including temporal distribution, packet length distribution, port-based distribution, protocol-based distribution, and TopN statistics. Such analysis not only provides support for the study of user behavior, but also enriches traffic measurement technology