Measurements and characterization of traffic in a university environment

ABSTRACT

In this paper we characterize the traffic seen in a university environment, especially on the broadband medium which serves as the backbone. The traffic is characterized based on parameters such as the volume of traffic generated, the packet inter-arrival times, the packet size and the protocol type and application that generated the traffic. We use the packet train model to analyze and characterize the traffic. Network traffic being inherently bursty, we propose a practical definition for burstiness and evaluate the burstiness of the traffic measured. Finally we compare the appropriateness and relative efficiency of some of the tools used to measure network traffic and compare their performance.

REFERENCES

• 1: Ramón Cáceres , Peter B. Danzig , Sugih Jamin , Danny J. Mitzel, Characteristics of wide-area TCP/IP conversations, Proceedings of the conference on Communications architecture & protocols, p.101-112, September 03-06, 1991, Zurich, Switzerland
• 2: Steven A. Heimlich, Traffic characterization of the NSFNET national backbone, Proceedings of the 1990 ACM SIGMETRICS conference on Measurement and modeling of computer systems, p.257-258, April 1990, Univ. of Colorado, Boulder, Colorado, United States
• 3: Jain Raj, and Routhier A Shawn, "Packet Trains - Measurements and a New Model for Computer Network Traffic", IEEE Journal on Selected Areas in Communication, Vol. SAC- 4, No. 6, Sept 1986.
• 4: Jungok, J and Suda T., "Survey of Traffic Control Schemes and Protocols in ArM Networks", Proceedings of the IEEE, Vol 79, No 2, Feb 1991.
• 5: Pawlita F. Peter, "Traffic Measurements in Data Networks, Recent Measurement Results, and Some Implications", IEEE Transactions on Communications, Vol. Com-29, No. 4, April 1981.
• 6: Network General Corporation, Sniffer Monitor, Sniffer Analyser Manuals, Series 500.
• 7: SUNOS SYSTEM, ETHERFIND Manual Pages, 1989
• 8: SUNOS SYSTEM, NFSwatch Manual Pages, 1991