Blind equalization using matrix momentum and natural gradient adaptation

G.H. Morison, T.S. Durrani

Research output: Contribution to conference › Paper

1 Citation (Scopus)

Abstract

In this paper the problem of single input multiple output (SIMO) and multiple input multiple output (MIMO) blind equalization in a frequency selective environment is addressed using blind source separation techniques. A robust whitening stage is included to reduce the effects of noise enhancement that traditional prewhitening methods suffer, and the use of matrix momentum with the natural gradient algorithm is utilised to improve the computation efficiency of the standard natural gradient algorithm. The performance of the algorithm is demonstrated for an ill conditioned channel and compared with a current natural gradient based blind equalization using source separation method.

LanguageEnglish
Pages439-448
Number of pages10
DOIs
Publication statusPublished - Sep 2003
Event2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03 - Toulouse, France
Duration: 17 Sep 2003 → 19 Sep 2003

Conference

Conference2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03
CountryFrance
CityToulouse
Period17/09/03 → 19/09/03

Fingerprint

Blind equalization
Momentum
Source separation
Blind source separation

Keywords

  • blind equalizers
  • source separation
  • sensor arrays
  • noise robustness
  • MIMO
  • frequency
  • filtering
  • equations
  • delay
  • blind source separation
  • blind equalization
  • matrix momentum
  • natural gradient adaptation

Cite this

Morison, G. H., & Durrani, T. S. (2003). Blind equalization using matrix momentum and natural gradient adaptation. 439-448 . Paper presented at 2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03, Toulouse, France. https://doi.org/10.1109/NNSP.2003.1318043
Morison, G.H. ; Durrani, T.S. / Blind equalization using matrix momentum and natural gradient adaptation. Paper presented at 2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03, Toulouse, France.10 p.
@conference{73bd1c5350074f5680a731db3bdd643d,
title = "Blind equalization using matrix momentum and natural gradient adaptation",
abstract = "In this paper the problem of single input multiple output (SIMO) and multiple input multiple output (MIMO) blind equalization in a frequency selective environment is addressed using blind source separation techniques. A robust whitening stage is included to reduce the effects of noise enhancement that traditional prewhitening methods suffer, and the use of matrix momentum with the natural gradient algorithm is utilised to improve the computation efficiency of the standard natural gradient algorithm. The performance of the algorithm is demonstrated for an ill conditioned channel and compared with a current natural gradient based blind equalization using source separation method.",
keywords = "blind equalizers , source separation , sensor arrays , noise robustness , MIMO, frequency, filtering, equations, delay, blind source separation, blind equalization, matrix momentum, natural gradient adaptation",
author = "G.H. Morison and T.S. Durrani",
year = "2003",
month = "9",
doi = "10.1109/NNSP.2003.1318043",
language = "English",
pages = "439--448",
note = "2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03 ; Conference date: 17-09-2003 Through 19-09-2003",

}

Morison, GH & Durrani, TS 2003, 'Blind equalization using matrix momentum and natural gradient adaptation' Paper presented at 2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03, Toulouse, France, 17/09/03 - 19/09/03, pp. 439-448 . https://doi.org/10.1109/NNSP.2003.1318043

Blind equalization using matrix momentum and natural gradient adaptation. / Morison, G.H.; Durrani, T.S.

2003. 439-448 Paper presented at 2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03, Toulouse, France.

Research output: Contribution to conference › Paper

TY - CONF

T1 - Blind equalization using matrix momentum and natural gradient adaptation

AU - Morison, G.H.

AU - Durrani, T.S.

PY - 2003/9

Y1 - 2003/9

N2 - In this paper the problem of single input multiple output (SIMO) and multiple input multiple output (MIMO) blind equalization in a frequency selective environment is addressed using blind source separation techniques. A robust whitening stage is included to reduce the effects of noise enhancement that traditional prewhitening methods suffer, and the use of matrix momentum with the natural gradient algorithm is utilised to improve the computation efficiency of the standard natural gradient algorithm. The performance of the algorithm is demonstrated for an ill conditioned channel and compared with a current natural gradient based blind equalization using source separation method.

AB - In this paper the problem of single input multiple output (SIMO) and multiple input multiple output (MIMO) blind equalization in a frequency selective environment is addressed using blind source separation techniques. A robust whitening stage is included to reduce the effects of noise enhancement that traditional prewhitening methods suffer, and the use of matrix momentum with the natural gradient algorithm is utilised to improve the computation efficiency of the standard natural gradient algorithm. The performance of the algorithm is demonstrated for an ill conditioned channel and compared with a current natural gradient based blind equalization using source separation method.

KW - blind equalizers

KW - source separation

KW - sensor arrays

KW - noise robustness

KW - MIMO

KW - frequency

KW - filtering

KW - equations

KW - delay

KW - blind source separation

KW - blind equalization

KW - matrix momentum

KW - natural gradient adaptation

UR - http://cogsys.imm.dtu.dk/nnsp2003/

U2 - 10.1109/NNSP.2003.1318043

DO - 10.1109/NNSP.2003.1318043

M3 - Paper

SP - 439

EP - 448

ER -

Morison GH, Durrani TS. Blind equalization using matrix momentum and natural gradient adaptation. 2003. Paper presented at 2003 IEEE 13th Workshop on Neural Networks for Signal Processing, 2003. NNSP'03, Toulouse, France. https://doi.org/10.1109/NNSP.2003.1318043

Access to Document