Neural Network Equalization for Asynchronous Multitrack Detection in TDMR

Elnaz Banan Sadeghian

doi:10.1109/TMRC56419.2022.9918163

Neural Network Equalization for Asynchronous Multitrack Detection in TDMR

Elnaz Banan Sadeghian

Department of Electrical and Computer Engineering

Stevens Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

The advent of multiple readers in magnetic recording opens the possibility of replacing the current industry's single-track detection with the more promising multitrack detection architectures. We have proposed a first solution, a generalized partial-response maximum-likelihood (GPRML) architecture, that extends the conventional PRML paradigm to jointly detect multiple asynchronous tracks. In this paper, we propose to replace the conventional communication-theoretic multiple-input multiple-output equalizer in the GPRML architecture with a neural network equalizer for better adaption to the nonlinearity of the underlying channel. We evaluate the proposed equalization strategy on a realistic two-dimensional magnetic-recording channel, and find that the proposed equalizer outperforms the conventional linear equalizer, by a 37% reduction in the bit-error rate and a 33% gain in the areal density.

Original language	English
Title of host publication	2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781665489065
DOIs	https://doi.org/10.1109/TMRC56419.2022.9918163
State	Published - 2022
Event	33rd IEEE Magnetic Recording Conference, TMRC 2022 - Milpitas, United States Duration: Aug 29 2022 → Aug 31 2022

Publication series

Name	2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings

Conference

Conference	33rd IEEE Magnetic Recording Conference, TMRC 2022
Country/Territory	United States
City	Milpitas
Period	8/29/22 → 8/31/22

ASJC Scopus subject areas

Media Technology
Electronic, Optical and Magnetic Materials
Instrumentation

Keywords

Intertrack interference
joint multiuser detection
multiple-input multiple-output (MIMO) channel
nonlinear equalization
timing recovery
two-dimensional magnetic recording (TDMR)

Access to Document

10.1109/TMRC56419.2022.9918163

Cite this

@inproceedings{13bec30aba324e418c6919df1a647c73,

title = "Neural Network Equalization for Asynchronous Multitrack Detection in TDMR",

abstract = "The advent of multiple readers in magnetic recording opens the possibility of replacing the current industry's single-track detection with the more promising multitrack detection architectures. We have proposed a first solution, a generalized partial-response maximum-likelihood (GPRML) architecture, that extends the conventional PRML paradigm to jointly detect multiple asynchronous tracks. In this paper, we propose to replace the conventional communication-theoretic multiple-input multiple-output equalizer in the GPRML architecture with a neural network equalizer for better adaption to the nonlinearity of the underlying channel. We evaluate the proposed equalization strategy on a realistic two-dimensional magnetic-recording channel, and find that the proposed equalizer outperforms the conventional linear equalizer, by a 37\% reduction in the bit-error rate and a 33\% gain in the areal density.",

keywords = "Intertrack interference, joint multiuser detection, multiple-input multiple-output (MIMO) channel, nonlinear equalization, timing recovery, two-dimensional magnetic recording (TDMR)",

author = "Sadeghian, \{Elnaz Banan\}",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 33rd IEEE Magnetic Recording Conference, TMRC 2022 ; Conference date: 29-08-2022 Through 31-08-2022",

year = "2022",

doi = "10.1109/TMRC56419.2022.9918163",

language = "English",

series = "2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings",

}

Sadeghian, EB 2022, Neural Network Equalization for Asynchronous Multitrack Detection in TDMR. in 2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings. 2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 33rd IEEE Magnetic Recording Conference, TMRC 2022, Milpitas, United States, 8/29/22. https://doi.org/10.1109/TMRC56419.2022.9918163

Neural Network Equalization for Asynchronous Multitrack Detection in TDMR. / Sadeghian, Elnaz Banan.
2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2022. (2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Neural Network Equalization for Asynchronous Multitrack Detection in TDMR

AU - Sadeghian, Elnaz Banan

PY - 2022

Y1 - 2022

N2 - The advent of multiple readers in magnetic recording opens the possibility of replacing the current industry's single-track detection with the more promising multitrack detection architectures. We have proposed a first solution, a generalized partial-response maximum-likelihood (GPRML) architecture, that extends the conventional PRML paradigm to jointly detect multiple asynchronous tracks. In this paper, we propose to replace the conventional communication-theoretic multiple-input multiple-output equalizer in the GPRML architecture with a neural network equalizer for better adaption to the nonlinearity of the underlying channel. We evaluate the proposed equalization strategy on a realistic two-dimensional magnetic-recording channel, and find that the proposed equalizer outperforms the conventional linear equalizer, by a 37% reduction in the bit-error rate and a 33% gain in the areal density.

AB - The advent of multiple readers in magnetic recording opens the possibility of replacing the current industry's single-track detection with the more promising multitrack detection architectures. We have proposed a first solution, a generalized partial-response maximum-likelihood (GPRML) architecture, that extends the conventional PRML paradigm to jointly detect multiple asynchronous tracks. In this paper, we propose to replace the conventional communication-theoretic multiple-input multiple-output equalizer in the GPRML architecture with a neural network equalizer for better adaption to the nonlinearity of the underlying channel. We evaluate the proposed equalization strategy on a realistic two-dimensional magnetic-recording channel, and find that the proposed equalizer outperforms the conventional linear equalizer, by a 37% reduction in the bit-error rate and a 33% gain in the areal density.

KW - Intertrack interference

KW - joint multiuser detection

KW - multiple-input multiple-output (MIMO) channel

KW - nonlinear equalization

KW - timing recovery

KW - two-dimensional magnetic recording (TDMR)

UR - https://www.scopus.com/pages/publications/85141653930

UR - https://www.scopus.com/pages/publications/85141653930#tab=citedBy

U2 - 10.1109/TMRC56419.2022.9918163

DO - 10.1109/TMRC56419.2022.9918163

M3 - Conference contribution

T3 - 2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings

BT - 2022 IEEE 33rd Magnetic Recording Conference, TMRC 2022 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 33rd IEEE Magnetic Recording Conference, TMRC 2022

Y2 - 29 August 2022 through 31 August 2022

ER -

Neural Network Equalization for Asynchronous Multitrack Detection in TDMR

Abstract

Publication series

Conference

ASJC Scopus subject areas

Keywords

Access to Document

Other files and links

Fingerprint

Cite this