Wide-area traffic: the failure of Poisson modeling
SIGCOMM '94 Proceedings of the conference on Communications architectures, protocols and applications
Statistical bandwidth sharing: a study of congestion at flow level
Proceedings of the 2001 conference on Applications, technologies, architectures, and protocols for computer communications
Insensitive Bandwidth Sharing in Data Networks
Queueing Systems: Theory and Applications
Evaluating the number of active flows in a scheduler realizing fair statistical bandwidth sharing
SIGMETRICS '05 Proceedings of the 2005 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
A queueing analysis of max-min fairness, proportional fairness and balanced fairness
Queueing Systems: Theory and Applications
The Erlang model with non-poisson call arrivals
SIGMETRICS '06/Performance '06 Proceedings of the joint international conference on Measurement and modeling of computer systems
QoS-aware bandwidth provisioning for IP network links
Computer Networks: The International Journal of Computer and Telecommunications Networking
Congestion in large balanced multirate links
Proceedings of the 23rd International Teletraffic Congress
| Hi-index | 0.00 |
We demonstrate that the Internet has a formula linking demand, capacity and performance that in many ways is the analogue of the Erlang loss formula of telephony. Surprisingly, this formula is none other than the Erlang delay formula. It provides an upper bound on the probability a flow of given peak rate suffers degradation when bandwidth sharing is max-min fair. Apart from the flow rate, the only relevant parameters are link capacity and overall demand. We explain why this result is valid under a very general and realistic traffic model and discuss its significance for network engineering.