Improved detection for magnetic recording systems with media noise

Caroselli, J.; Altekar, S.A.; McEwen, P.A.; Wolf, J.K.

doi:10.1109/20.617728

Cited by 31 publications

(9 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Media noise, which is nonstationary, correlated, and data-dependent, can cause serious performance degradation in a TDMR system when compared to a system operating with the same amount of AWGN. Correspondingly, a data-dependent full local feedback noise prediction scheme [Caroselli 1997] has been adopted for the RSH in our TDMR system. In this scheme, the Viterbi detector recursively updates the cumulative metrics for each state using the update equations…”

Section: B Detection With Rotated Single Head (Rsh)mentioning

confidence: 99%

Novel system design for readback at 10 terabits per square inch user areal density

2012

View full text Add to dashboard Cite

A novel system design for sensing very high density magnetic recording data, such as that envisioned for twodimensional magnetic recording, is proposed. The key idea is a rotated sense head, so that the shields are aligned downtrack, combined with oversampled signal processing to regain the lost down-track resolution. Based on a random Voronoi grain model, simulation indicates that for bits with dimension of 8 nm × 6 nm, 5.5 nm grains, and a reader with 4 nm × 18 nm × 18 nm free layer and 11 nm shield-shield spacing, the bit error rate can drop from 20.9% for a normally positioned head array to 4.2% for a single-rotated single head with sampling period of 2 nm, a minimum mean squared error equalizer, and pattern-dependent noise prediction detector. The user density computed using the Shannon capacity limit is greatly increased to 10.1 Tb/in 2 .

show abstract

Section: B Detection With Rotated Single Head (Rsh)mentioning

confidence: 99%

Novel system design for readback at 10 terabits per square inch user areal density

2012

View full text Add to dashboard Cite

show abstract

“…We should also emphasize that the predictor and its error variance depend on the bit pattern . We now write (4) where is optimal linear prediction of and is the predictor error variance. As will be shown later, using the medium noise model of our interest, optimal predictors can be approximated by finite length predictors of order , which use most recent noise samples.…”

Section: A Derivationmentioning

confidence: 99%

“…So the branch metric reduces to one corresponding to uncorrelated noise with signal-dependent variance. Comparing the branch metric of (11) with that used in [4], we see that the approach of [4] ignores the branch-dependent noise variance term in the metric computation. It can also be verified that with an appropriate model parameter selection, the branch metric derived in [6] reduces to (11).…”

Section: A Derivationmentioning

confidence: 99%

“…The performance improvement comes from the effective whitening of the noise samples that became correlated at the detector input due to the equalization constraint [1]- [3]. This approach, called the noise predictive maximum likelihood (NPML) method, also has been extended to signal-dependent medium noise channels, where the noise characteristics depend highly on the local bit patterns [4]. This paper is intended to provide a more formal and general treatment of the NPML as applied to channels subject to signal-dependent noise.…”

Section: Introductionmentioning

confidence: 99%

“…This provides significant insight into other suboptimal detector structures as well as performance analysis. In particular, we note that the approach of [4] is a special suboptimal realization of a broader class of detectors. Instead of forcing an AR noise process with arbitrarily specified model order parameters as in [5] and [6], we use the established noise model that arises from physical interpretation of the medium noise process.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Pattern-dependent noise prediction in signal-dependent noise

Moon

Park

2001

IEEE J. Select. Areas Commun.

112

View full text Add to dashboard Cite

Maximum and near-maximum likelihood sequence detectors in signal-dependent noise are discussed. It is shown that the linear prediction viewpoint allows a very simple derivation of the branch metric expression that has previously been shown as optimum for signal-dependent Markov noise. The resulting detector architecture is viewed as a noise predictive maximum likelihood detector that operates on an expanded trellis and relies on computation of branch-specific, pattern-dependent noise predictor taps and predictor error variances. Comparison is made on the performance of various low-complexity structures using the position-jitter/width-variation model for transition noise. It is shown that when medium noise dominates, a reasonably low complexity detector that incorporates pattern-dependent noise prediction achieves a significant signal-to-noise ratio gain relative to the extended class 4 partial response maximum likelihood detector. Soft-output detectors as well as the use of soft decision feedback are discussed in the context of signal-dependent noise.Index Terms-Markov process, maximum likelihood sequence detection, noise prediction, reduced-complexity sequence detector, signal-dependent noise.

show abstract

Signal Processing for Magnetic‐Recording Channels

Eleftheriou

2003

Wiley Encyclopedia of Telecommunications

View full text Add to dashboard Cite

In the past decade, several advanced digital signal‐processing and coding techniques have been introduced into hard‐disk drives to improve the error‐rate performance at ever increasing normalized linear densities as well as to reduce manufacturing and servicing costs. The state of the art in signal processing and constrained coding for hard‐disk drives is reviewed, with emphasis on the techniques that have been used in commercial systems.

show abstract

Improved detection for magnetic recording systems with media noise

Cited by 31 publications

References 3 publications

Novel system design for readback at 10 terabits per square inch user areal density

Novel system design for readback at 10 terabits per square inch user areal density

Pattern-dependent noise prediction in signal-dependent noise

Signal Processing for Magnetic‐Recording Channels

Contact Info

Product

Resources

About