Skip to Main content Skip to Navigation
Conference papers

Erlang-based dimensioning for IPv4 Address+Port translation

Abstract : As the IPv4 address pool is being exhausted, it becomes urgent to find a way to migrate IPv4 network architectures to IPv6, or to reduce the use of IPv4 addresses. In this paper, we discuss a strategy known as ''Address + Port'' translation, which consists in several users sharing the same IPv4 address and being distinguished by a range of port numbers. Of critical importance for the feasibility of such a mechanism is the knowledge of the minimum number of ports to allocate to users so that no service degradation is perceived. To that extent, we analyse the port consumption of the most port-consuming Internet applications, web browsing, and present some aggregate port consumption curves for the student population of our campus. Our results suggest that a port range of 1000 ports is totally transparent to users (which would allow to share a single IPv4 address among 64 users),while 400 ports (i.e., 150 users per address) is sufficient for most of users. Finally, the number of users per address could be further improved by benefiting from statistical multiplexing, i.e., using dynamical instead of fixed port range allocation.
Document type :
Conference papers
Complete list of metadata

Cited literature [5 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00785727
Contributor : Bibliothèque Télécom Bretagne <>
Submitted on : Wednesday, February 6, 2013 - 6:38:30 PM
Last modification on : Tuesday, April 20, 2021 - 10:32:06 AM
Long-term archiving on: : Tuesday, May 7, 2013 - 5:35:14 AM

File

erlang-based_dimensioning_for_...
Files produced by the author(s)

Identifiers

Citation

Florent Fourcot, Bertrand Grelot, Isabelle Kraemer, Frédéric Perrin, Patrick Maillé, et al.. Erlang-based dimensioning for IPv4 Address+Port translation. ICC 2012: IEEE International Conference on Communications, Jun 2012, Ottawa, Canada. pp.1230-1234, ⟨10.1109/ICC.2012.6364328⟩. ⟨hal-00785727⟩

Share

Metrics

Record views

1229

Files downloads

2036