Back to results
US 11,080,621 B2Granted
Machine learning-based read channel data detection
AssigneeWestern Digital Technologies, Inc.
InventorsMichael Alex, Xinzhi Xing
Life of this patent
Estimating the term…
Abstract
Technology for improved data detection using machine learning may include a method in which an analog read signal comprising data read from a non-transitory storage medium of the data storage device is received. The analog read signal is processed into a plurality of digital samples. A digital sample from the plurality of digital samples is classified into a category from a plurality of categories using a machine learning algorithm for at least some of the plurality of digital samples. The plurality of digital samples is then decoded based on at least some of the predicted categories.
Claims23
A read channel circuit, comprising:
- an analog to digital converter (ADC) configured to receive an analog read signal and convert the analog read signal into a plurality of digital samples;
- a machine learning detector coupled to the ADC, the machine learning detector being configured to receive the plurality of digital samples from the ADC and predict to which category, from a plurality of categories, a digital sample from the plurality of digital samples belongs, for at least some of the plurality of digital samples;
- and
- a decoder configured to decode, based on at least some of the predicted categories, the plurality of digital samples to reveal previously encoded data.
1 more element
A reception circuit of a transceiver, comprising:
- an analog to digital converter (ADC) configured to receive an analog input signal and convert the analog input signal into a plurality of digital samples;
- a machine learning detector coupled to the ADC, the machine learning detector being configured to receive the plurality of digital samples from the ADC and predict to which category, from a plurality of categories, a digital sample from the plurality of digital samples belongs, for at least some of the plurality of digital samples;
- and
- a decoder configured to decode, based on at least some of the predicted categories, the plurality of digital samples to reveal previously encoded data.
1 more element
A data storage device comprising:
- a non-transitory storage medium;
- and
- control circuitry configured to:
- receive an input signal comprising data read from the non-transitory storage medium, the input signal including a non-linear and distorted waveform;
- process the input signal into a plurality of digital samples;
- classify, using a machine learning algorithm, a digital sample from the plurality of digital samples into a category from a plurality of categories for at least some of the plurality of digital samples;
- decode the plurality of digital samples based on at least some of the plurality of categories;
- and
- descramble the decoded plurality of digital samples to reveal the data read from non-transitory storage medium.
6 more elements
A method comprising:
- receiving an analog read signal comprising data read from a non-transitory storage medium of a non-transitory data storage device;
- processing the analog read signal into a plurality of digital samples;
- classifying, using a machine learning algorithm, a digital sample from the plurality of digital samples into a category from a plurality of categories for at least some of the plurality of digital samples;
- and
- decoding, based on at least some of the plurality of categories, the plurality of digital samples to reveal previously encoded data.
2 more elements
A method comprising:
- receiving a data transmission as an input signal;
- processing the input signal into a plurality of digital samples;
- classifying, using a machine learning algorithm, a digital sample from the plurality of digital samples into a category from a plurality of categories for at least some of the plurality of digital samples;
- and
- decoding, based on at least some of the plurality of categories, the plurality of digital samples to reveal previously encoded data.
2 more elements
A read channel circuit, comprising:
- an analog to digital converter (ADC) configured to receive an analog read signal and convert the analog read signal into a plurality of digital samples;
- a machine learning detector coupled to the ADC, the machine learning detector being configured to receive the plurality of digital samples from the ADC and predict to which category, from a plurality of categories, a digital sample from the plurality of digital samples belongs, for at least some of the plurality of digital samples;
- an electronic finite impulse response (FIR) filter situated between the ADC and the machine learning detector, the FIR filter configured to receive the plurality of digital samples and process the plurality of digital samples for input into the machine learning detector;
- and
- a decoder configured to decode the plurality of digital samples based on at least some of the predicted categories.
2 more elements
A read channel circuit, comprising:
- an analog to digital converter (ADC) configured to receive an analog read signal and convert the analog read signal into a plurality of digital samples;
- a machine learning detector coupled to the ADC, wherein:
- the machine learning detector is configured to receive the plurality of digital samples from the ADC and predict to which category, from a plurality of categories, a digital sample from the plurality of digital samples belongs, for at least some of the plurality of digital samples;
- the machine learning detector comprises a plurality of detectors that are configured to process each digital sample independently in parallel;
- and
- each detector from the plurality of detectors is configured to output a probability value for each digital sample reflecting a probability of that digital sample belonging to a category;
- and
- a decoder configured to decode the plurality of digital samples based on at least some of the predicted categories.
5 more elements
Dependencies are inferred from the claim text.