Adaptive Data Detection on Nonlinear Magnetic Tape Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic tape storage systems face significant challenges due to nonlinear effects such as nonlinear transition shifts and nonlinearities in magneto-resistive read transducers, which affect data recovery and cannot be entirely eliminated by existing noise-predictive detection methods.
Innovation Solution
A data storage system that includes a head for producing signals from a storage medium, an estimator to determine both linear and nonlinear portions of the signal, noise whitening filters to process the difference between the signal and the estimated signal, and an adaptive data-dependent noise-predictive maximum likelihood sequence detector to generate output streams based on branch metrics, effectively reducing nonlinearity and improving signal fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If noise-predictive detection is used with linear estimate of PR4 signal, then detection performance is improved, but nonlinear effects (NLTS and read transducer nonlinearity) cause bit error rates to increase
Solution Approach 1:
The patent segments the signal into linear and nonlinear portions by using an estimator to separate the PR4 equalizer output into these components. This allows the detection system to process the linear portion while compensating for the nonlinear portion, thereby improving detection performance without suffering from the bit error rate increase caused by nonlinear effects.
Solution Approach 2:
The patent changes the parameter representation by modeling the nonlinear channel as a time-varying linear channel with state-dependent parameters. The estimator dynamically adjusts the linear estimate parameters based on the current state, allowing the system to maintain high detection performance while compensating for nonlinear effects that would otherwise increase bit error rates.
2Manufacturing precision
If write compensation is applied to reduce nonlinear transition shifts, then manufacturing precision is improved, but nonlinear effects cannot be completely eliminated
Solution Approach 1:
The patent implements feedback through the estimator that continuously monitors the signal and updates the linear estimate parameters. This feedback mechanism compensates for residual nonlinear effects that remain after write compensation, thereby maintaining high manufacturing precision while eliminating the reliability issues caused by residual nonlinear effects.
Solution Approach 2:
The patent uses a composite approach by combining write compensation techniques with adaptive noise-predictive detection that models the channel as time-varying linear. This composite strategy achieves both improved manufacturing precision and eliminated residual nonlinear effects, resolving the contradiction between the two parameters.
3Device complexity
If linear estimate of PR4 signal is used, then device complexity is reduced, but measurement precision deteriorates due to unaccounted nonlinearities
Solution Approach 1:
The patent introduces dynamics by making the linear estimate parameters time-varying and state-dependent rather than fixed. The estimator dynamically adapts the parameters based on the current channel state, which maintains low device complexity while significantly improving signal estimation accuracy by accounting for nonlinearities through the time-varying parameter model.
4Reliability
If adaptive estimation of nonlinear portion is implemented, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent applies preliminary action by having the estimator pre-process the signal to separate linear and nonlinear portions before the main detection process. This preliminary estimation improves reliability by accounting for nonlinearities early in the signal processing chain, while keeping device complexity manageable by performing the estimation in a dedicated preprocessing stage rather than throughout the entire detection system.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The proposed system reduces bit error rates and increases linear density by canceling deterministic signal nonlinearity before noise prediction, leading to more accurate data recovery and reduced errors in error correction codes.
Implementation Method 1
a head configured to produce a signal representing data stored on a storage medium
Implementation Method 2
a head configured to produce a signal representing data stored on a storage medium
Implementation Method 3
an estimator configured to determine an estimated signal from the signal, the estimated signal comprising a superposition of an estimated linear portion of a partial-response equalizer output and an estimated nonlinear portion of the signal
Implementation Method 4
a bank of noise whitening filters configured to apply one or more noise whitening filters to a difference between the signal and the estimated signal to produce a filtered signal
Implementation Method 5
a branch metric calculator configured to perform one or more branch metric calculations on a metric input signal based on the filtered signal to generate one or more branch metrics
Implementation Method 6
an adaptive data-dependent noise-predictive maximum likelihood sequence detector configured to generate an output stream representing the data based on the one or more branch metrics
Data Source
AI summary
A data storage system comprises: a head configured to produce a signal representing data stored on a storage medium; an estimator configured to determine an estimated signal comprising a superposition of an estimated linear portion of a partial-response equalizer output and an estimated nonlinear portion of the signal; a bank of noise whitening filters configured for filtering a difference between the signal and the estimated signal; a branch metric calculator configured to calculate branch metrics based on the filtered signal; and an adaptive data-dependent noise-predictive maximum likelihood sequence detector configured to generate an output stream representing the data based on the one or more branch metrics.