Statistical multiplexing over DSL networks

Jianwei Huang, Chee Wei Tan, Mung Chiang, Raphael Cendrillon

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations


Most previous work in statistical multiplexing only considered the case where the link transmission rates are fixed. In this paper, we consider statistical multiplexing in networks with adaptive transmission rates, with focus on DSL broadband access networks. This requires a jointly optimized allocation of buffer space and transmission bandwidth to traffic flows, which takes the flow traffic characteristics, the user QoS requirements, and the user interactions at the physical layer into consideration. Using the Effective Bandwidth concept, we propose a class of Alternate Maximization (AM) algorithms (AM-D and AM-M), which solve the statistical multiplexing problem for both delay insensitive Data traffic and delay sensitive Multimedia traffic. With low complexity as a design goal, the AM algorithms incorporate our recently proposed Autonomous Spectrum Balancing (ASB) algorithm, which was originally designed for DSL physical layer spectrum management. Our numerical results show that the AM algorithms combines the gain due to statistical multiplexing and that due to spectrum management.

Original languageEnglish (US)
Title of host publicationProceedings - IEEE INFOCOM 2007
Subtitle of host publication26th IEEE International Conference on Computer Communications
Number of pages9
StatePublished - 2007
EventIEEE INFOCOM 2007: 26th IEEE International Conference on Computer Communications - Anchorage, AK, United States
Duration: May 6 2007May 12 2007

Publication series

NameProceedings - IEEE INFOCOM
ISSN (Print)0743-166X


OtherIEEE INFOCOM 2007: 26th IEEE International Conference on Computer Communications
Country/TerritoryUnited States
CityAnchorage, AK

All Science Journal Classification (ASJC) codes

  • General Computer Science
  • Electrical and Electronic Engineering


Dive into the research topics of 'Statistical multiplexing over DSL networks'. Together they form a unique fingerprint.

Cite this