Cooperative Sequence Equalization for Band-Limited Channel ISI

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In magnetic storage systems, increasing data density leads to intersymbol interference (ISI) due to overlapping magnetic pulses, causing detection errors in band-limited channels.

Innovation Solution

A method involving cooperative sequence equalization adaptation techniques, where a sample stream is input to equalization filter banks and noise predictive filters to generate filtered equalization streams, reducing noise variance without compromising Viterbi-based detection processes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If data density is increased to improve storage capacity, then storage capacity is improved, but intersymbol interference increases causing detection errors

Engineering Contradiction:

Improvestorage capacityVSAvoiddetection accuracy

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

The patent divides the equalization process into multiple filter banks (first, second, third filter banks) with different filter types. Each filter bank processes the sample stream independently to produce separate equalized streams, which are then combined. This segmentation allows the system to handle ISI more effectively while maintaining detection accuracy even at increased data densities.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent combines multiple equalized streams from different filter banks through a combining operation to produce a final equalized stream. This merging of multiple processing paths enhances the overall equalization performance, reducing ISI effects and improving detection reliability without sacrificing storage capacity.

Inventive Principle:

Principle #5Merging (Combining)

2Reliability

If traditional equalization methods are used to reduce ISI, then detection errors are reduced, but noise variance remains high degrading detection accuracy

Engineering Contradiction:

Improvedetection accuracyVSAvoidnoise variance

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent applies different filter types (first, second, and third filter banks with distinct filter characteristics) to different portions of the signal processing. Each filter bank is optimized for specific signal characteristics, allowing local optimization of noise reduction while maintaining overall detection accuracy. This local quality approach enables targeted noise variance reduction without compromising the equalization of ISI-affected signals.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS8917470B2Data sequence detection in band-limited channels using cooperative sequence equalization

Publication Date: 2014.12.23 AVAGO TECHNOLOGIES INTERNATIONAL SALES PTE LTD

AI summary

A method for detecting a data sequence includes generating a sample stream, which is a time-sequenced digital signal associated with samples of an analog signal. The sample stream is input to n equalization filter banks, which each have m equalization filters to generate m equalized sample streams. The m equalized sample streams from each equalization filter bank are input to a corresponding one of n noise predictive filters. Each noise predictive filter is an m-tap noise predictive filter that receives the m equalized sample streams from a corresponding one of the n equalization filter banks. Each noise predictive filter is associated with one of n data patterns. A filtered equalization stream is generated by each noise predictive filter. Noise sample streams are generated using the filtered equalization streams generated by the n noise predictive filters. A data sequence is detected using the noise sample streams.