Impulsive noise estimation and cancellation in DSL using compressive sampling

T. Y. Al-Naffouri*, F. F. Al-Shaalan, A. A. Quadeer, H. Hmida

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

24 Scopus citations

Abstract

Impulsive noise is the bottleneck that determines the maximum length of the DSL. Impulsive noise seldom occurs in DSL but when it occurs, it is very destructive and results in dropping the affected DSL symbols at the receiver as they cannot be recovered. By considering impulsive noise a sparse vector, recently developed sparse reconstruction algorithms can be utilized to combat it. We propose an algorithm that utilizes the null carriers for the impulsive noise estimation and cancellation. Specifically, we use compressive sampling for a coarse estimate of the impulse position, an a priori information based MAP metric for its refinement, followed by MMSE estimation for estimating the impulse amplitudes. We also present a comparison of the achievable rate in DSL using our algorithm and recently developed algorithms for sparse signal reconstruction.

Original languageEnglish (US)
Title of host publication2011 IEEE International Symposium of Circuits and Systems, ISCAS 2011
Pages2133-2136
Number of pages4
DOIs
StatePublished - 2011
Externally publishedYes
Event2011 IEEE International Symposium of Circuits and Systems, ISCAS 2011 - Rio de Janeiro, Brazil
Duration: May 15 2011May 18 2011

Publication series

NameProceedings - IEEE International Symposium on Circuits and Systems
ISSN (Print)0271-4310

Other

Other2011 IEEE International Symposium of Circuits and Systems, ISCAS 2011
Country/TerritoryBrazil
CityRio de Janeiro
Period05/15/1105/18/11

Keywords

  • Compressive sampling
  • DSL
  • Impulsive noise
  • Sparse signal reconstruction

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Impulsive noise estimation and cancellation in DSL using compressive sampling'. Together they form a unique fingerprint.

Cite this