Identifying a state of a data storage drive using an artificial neural network generated model

• US 7,328,197 B2
• Filed: 09/23/2004
• Issued: 02/05/2008
• Est. Priority Date: 09/23/2004
• Status: Expired due to Fees

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Seger, Paul J
• Primary Examiner(s): STARKS, WILBERT L
• Application Number:
  US10/947,692
• Publication Number:
  US 20060074820A1
• Time in Patent Office: 1,230 Days
• Field of Search: 706/15, 706/45, 706/47, 707/102, 707/104.1, 702/184, 382/128, 735/99, 436/172
• US Class Current: 706/15
• CPC Class Codes: G06F 11/2263 using neural networks
• CPC Class Codes: G06F 3/0634 by changing the state or mode of one or more devices
• CPC Class Codes: G06N 3/02 Neural networks

Abstract

The state or condition of a data storage drive, or a subsystem within a drive, may be evaluated by comparing a set of selected parameter values, converted into a trial vector, with a number of model or exemplar vectors, each of which was represents a particular state or condition of a sample drive. Examples of such conditions may include “good”, “marginal”, “unacceptable”, “worn”, “defective”, or other general or specific conditions. Sets of parameter values from the drive are converted into input vectors. Unprocessed vectors are then processed against the input vectors in an artificial neural network to generate the exemplar vectors. The exemplar vectors are stored in a memory of an operational drive. During operation of the drive, the trial vector is compared with the exemplar vectors. The exemplar vector which is closest to the trial vector represents a state which most closely represents the current state of the drive. Thus, a high similarity between the trial vector and an exemplar vector which represent a “good” drive is likely to have come from a “good” drive.

Claims

1. A data storage drive comprising:

a transducer adapted to read transitions representing data from a storage media and convert the transitions to an analog signal;

a sampler adapted to convert the analog signal into a digital format;

a detector adapted to identify transitions in the digital signal and a decoder adapted to generate an output substantially corresponding to the data;

a memory operable to store a plurality J of n-tuple exemplar vectors representative of potential states of the data storage drive;

the memory further operable to store a first predetermined value;

an interface coupled to receive a plurality n of current parameter values from the data storage drive;

means for generating an n-tuple trial vector from the n parameter values;

a processor operable to determine a distance between the trial vector and each exemplar vector;

the processor further operable to identify an exemplar vector K which is the least distance to the trial vector, the least distance comprising an activation value; and

a comparator operable to compare the activation value with the first predetermined value, whereby if the activation value is less than the first predetermined value, the current state of the data storage drive is characterized by the exemplar vector K.

2. The storage drive of claim 1, the means for generating the trial vector further operable to normalize the trial vector to a predetermined magnitude.

3. The storage drive of claim 1, wherein the processor is further operable to calculate the Euclidean distance between each exemplar vector and the trial vector.

4. The storage drive of claim 1, whereby, if the activation value is greater than the first predetermined value, the trial vector is invalid.

5. The storage drive of claim 1, wherein the comparator is further operable to:

compare the activation value with a second predetermined value, the second predetermined value being less than the first predetermined value, and if the activation value is less than the second predetermined value, log the activation value and if the activation value is greater than the second predetermined value, initiate a system calibration; and

compare the activation value with a third predetermined value, the third predetermined value being greater than the second predetermined value and less than the first predetermined value, and if the activation value is greater than the third predetermined value, set an alarm.

6. The storage drive of claim 1, further comprising an equalizer coupled between the sampler and the detector, the equalizer comprising an adaptive FIR filter having a plurality T of taps of which n taps are coupled to the interface, wherein the potential states represented by the exemplar vectors comprise potential states of the FIR taps.

7. The storage drive of claim 6, wherein the processor is further operable to convert the values of the FIR taps from the time domain to the frequency domain to substantially remove ambiguities associated with the FIR taps.

8. The storage drive of claim 6, wherein the processor is further operable to select the n taps to couple to the interface, where n<

T.

9. A method for evaluating the state of an adaptive equalizer in a data storage drive read/write channel, comprising:

generating a plurality J of n-tuple exemplar vectors representative of potential states of the equalizer;

operating the read/write channel;

generating an n-tuple trial vector from n FIR tap values representing an actual state of the equalizer;

of the J exemplar vectors, identifying an exemplar vector K which is closest to the trial vector, the distance from the trial vector to the identified exemplar vector K being an activation value;

whereby, if the activation value is less than a first predetermined value, the actual state of the equalizer is characterized by the exemplar vector K.

10. The method of claim 9, wherein generating the exemplar vectors comprises:

initializing a plurality J of n-tuple unprocessed exemplar vectors with random values;

globally conditioning the J unprocessed exemplar vectors;

separating the conditioned exemplars; and

consolidating the separated exemplar vectors.

11. The method of claim 10, wherein globally conditioning the J unprocessed exemplar vectors comprises:

sequentially applying a plurality M of input vectors to the J unprocessed exemplar vectors;

updating the J unprocessed exemplar vectors by a factor of α

₁; and

repeating the applying and updating steps over a predetermined number e₁ of epochs.

12. The method of claim 10, wherein separating the conditioned vectors comprises:

sequentially applying the plurality M of input vectors to the J conditioned exemplar vectors;

of the J conditioned exemplar vectors, selecting the conditioned exemplar vector which is closest to the currently applied input vector;

updating the selected conditioned exemplar vector and a predetermined number of adjoining conditioned exemplar vectors by a factor of α

₂; and

repeating the applying, selecting and updating steps over a predetermined number e₂ of epochs.

13. The method of claim 10, wherein consolidating the separated vectors comprises:

sequentially applying the plurality M of input vectors to the J separated exemplar vectors;

of the J separated exemplar vectors, selecting the separated exemplar vector which is closest to the currently applied input vector;

updating the selected conditioned exemplar vector by a factor of α

₃; and

repeating the applying, selecting and updating steps over a predetermined number e₃ of epochs.

14. The method of claim 9, further comprising normalizing each unprocessed exemplar vector and the trial vector to a predetermined magnitude.

15. The method of claim 9, wherein identifying the exemplar vector K which is closest to the trial vector comprises calculating the Euclidean distance between each exemplar vector and the trial vector.

16. The method of claim 9, whereby, if the activation value is greater than the first predetermined value, the trial vector is invalid.

17. The method of claim 9, further comprising:

if the activation value is less than a second predetermined value, the second predetermined value being less than the first predetermined value, logging the activation value;

if the activation value is greater than the second predetermined value, initiating a read/write channel calibration; and

if the activation value is greater than a third predetermined value, the third predetermined value being greater than the second predetermined value and less than the first predetermined value, setting an alarm.

18. The method of claim 9, further comprising converting the FIR tap values from the time domain to the frequency domain to substantially remove ambiguities associated with the FIR taps.

19. The method of claim 9, wherein the equalizer comprises a plurality T of FIR tap values, the method further comprising selecting the n taps to couple to the interface, where n<

T.

20. A computer program product of a computer readable medium usable with a programmable computer, the computer program product having computer-readable code embodied therein for evaluating the state of a data storage drive read/write channel, the computer-readable code comprising instructions for:

operating the data storage drive;

generating an n-tuple trial vector from n parameters representing an actual state of the data storage drive;

retrieving a plurality of n-tuple exemplar vectors from a memory, each exemplar vector representing a potential state of the data storage drive; and

of the J exemplar vectors, identifying an exemplar vector K which is closest to the trial vector, the distance from the trial vector to the identified exemplar vector K being an activation value;

whereby, if the activation value is less than a first predetermined value the actual state of the data storage drive is characterized by the exemplar vector K.

21. The computer program product of claim 20, wherein the instructions further comprise instruction for normalizing the trial vector to a predetermined magnitude.

22. The computer program product of claim 20, wherein the instructions for determining which exemplar vector K is closest to the trial vector comprise instructions for calculating the Euclidean distance between each exemplar vector and the trial vector.

23. The computer program product of claim 20, whereby, if the activation value is greater than the first predetermined value, the trial vector is invalid.

24. The computer program product of claim 20, wherein the instructions further comprise instructions for:

logging the activation value if the activation value is less than a second predetermined value, the second predetermined value being less than the first predetermined value;

initiating a system calibration if the activation value is greater than the second predetermined value; and

setting an alarm if the activation value is greater than a third predetermined value, the third predetermined value being greater than the second predetermined value and less than the first predetermined value.

25. The computer program product of claim 20, wherein the instructions for generating an n-tuple trial vector further comprise instructions for obtaining the values of n taps from an adaptive FIR filter in an equalizer in the data storage drive and wherein the potential states represented by the exemplar vectors comprise potential states of the FIR taps.

26. The computer program product of claim 25, wherein the equalizer comprises a plurality T of FIR tap values and the instructions for generating an n-tuple trial vector further comprise instructions for selecting the n taps, where n<

T.

27. The computer program product of claim 25, wherein the computer-readable code further comprise instructions for converting the FIR tap values from the time domain to the frequency domain to substantially remove ambiguities associated with the FIR taps.