Systems methods and computer program products for controlling undesirable bias in an equalizer

• US 6,505,222 B1
• Filed: 10/29/1999
• Issued: 01/07/2003
• Est. Priority Date: 10/29/1999
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Ware, Malcolm Scott, Davis, Gordon Taylor
• Primary Examiner(s): NGO, CHUONG D
• Application Number:
  US09/430,578
• Time in Patent Office: 1,166 Days
• Field of Search: 708/300, 708/322, 708/323, 708/229, 708/230, 708/231, 708/232, 708/233
• US Class Current: 708/323
• CPC Class Codes: H03M 13/00 Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes error detection or error correction for analogue/digital, digital/analogue or code conversion H03M1/00 – H03M11/00; specially adapted for digital computers G06F11/08; for information storage based on relative movement between record carrier and transducer G11B, e.g. G11B20/18; for static stores G11C

Abstract

Bias control systems, methods, and computer program products generate an error signal that corresponds to a difference between a reference signal and an equalizer output signal. The error signal is then filtered using a first filter circuit to generate an error signal average. If the absolute value of the error signal does not exceed a suitable threshold that is proportional to the error signal average, then the error signal is coupled to the equalizer for use in updating the filter coefficients. Furthermore, a second filter circuit may be used to generate an average of selected equalizer output signal samples. If the absolute value of the error signal is greater than the threshold that is proportional to the error signal average, then the reference signal is updated to correspond to the average of selected equalizer output signal samples. Large errors are interpreted as resulting from inaccurate reference signals or reference levels. The reference signals or reference levels may be inaccurate due to the effects of digital impairments in the network, such as RBS, exhibited in the equalizer output signal. Rather than using the error signal to update the equalizer filter coefficients, the reference signal or reference level is updated to correspond to the average of selected equalizer output signal samples. Conversely, small errors are interpreted as an indication that the reference signals or reference levels are accurate and do not require additional refinement. In this case, the error signal is used to update the equalizer filter coefficients.

Claims

1. A bias control system for an equalizer, comprising:

an adder that is responsive to a reference signal and an equalizer output signal and generates an error signal that corresponds to a difference between the reference signal and the equalizer output signal;

a first filter circuit that is responsive to the error signal and generates an error signal average; and

a comparator that is responsive to the error signal and the error signal average and that allows the error signal to propagate to the equalizer for use in coefficient updating if the error signal average is at least as great as an absolute value of the error signal.

2. A bias control system as recited in claim 1, wherein the comparator generates a coefficient update signal to allow the error signal to propagate to the equalizer and the bias control system further comprises:

a gate that couples the error signal to the equalizer in response to the coefficient update signal.

3. A bias control system as recited in claim 1, further comprising:

a second filter that generates an average of selected equalizer output signal samples; and

wherein the comparator allows the reference signal to be updated to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

4. A bias control system as recited in claim 3, wherein the comparator generates a reference update signal and the bias control system further comprises:

a reference level memory array that contains a magnitude of the reference signal for each of a plurality of data frame intervals, the second filter circuit being responsive to the reference level update signal to update the magnitude of the reference signal for one of the plurality of data frame intervals to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

5. A bias control system as recited in claim 4, further comprising a data frame interval counter that generates a data frame interval output signal that identifies the one of the plurality of data frame intervals.

6. A bias control system as recited in claim 1, further comprising a scaling circuit that scales the error signal and provides the scaled error signal to the first filter circuit.

7. A method of controlling bias in an equalizer, comprising the steps of:

generating an error signal that corresponds to a difference between a reference signal and an equalizer output signal;

generating an error signal average; and

coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as an absolute value of the error signal.

8. A method as recited in claim 7, further comprising the steps of:

generating an average of selected equalizer output signal samples; and

updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

9. A method as recited in claim 8, wherein the updating step comprises the step of:

updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average and the equalizer is in a decision-directed mode.

10. A method as recited in claim 8, wherein the updating step comprises the step of:

updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average and the error signal average is less than an error convergence threshold.

11. A method as recited in claim 8, further comprising the step of:

storing a magnitude of the reference signal for each of a plurality of data frame intervals.

12. A method as recited in claim 11, wherein the updating step comprises the step of:

updating the magnitude of the reference signal for one of the plurality of data frame intervals to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

13. A method as recited in claim 11, wherein the step of generating the average of selected equalizer output signal samples comprises the steps of:

computing a first component by multiplying the magnitude of the reference signal for one of the plurality of data frame intervals by a first weight factor (WF1);

computing a second component by multiplying the equalizer output signal by a second weight factor (WF2); and

adding the first component to the second component to generate the average of selected equalizer output signal samples.

14. A method as recited in claim 13, wherein:

the WF1 is between zero and one;

the WF2 is between zero and one; and

the WF1 plus the WF2 is about one.

15. A method as recited in claim 13, wherein the WF1 is about {fraction (63/64)} and the WF2 is about {fraction (1/64)}.

16. A method as recited in claim 7, wherein the step of generating the error signal average comprises the steps of:

computing a first error signal average component by multiplying a previous error signal average by a first weight factor (WF1);

computing a second error signal average component by multiplying the absolute value of the error signal by a second weight factor (WF2); and

adding the first error signal average component to the second error signal average component to generate the error signal average.

17. A method as recited in claim 16, wherein:

the WF1 is between zero and one;

the WF2 is between zero and one; and

the WF1 plus the WF2 is about one.

18. A method as recited in claim 16, wherein the WF1 is about {fraction (31/32)} and the WF2 is about {fraction (1/32)}.

19. A method as recited in claim 7, wherein the coupling step comprises the steps of:

coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as the absolute value of the error signal;

coupling the error signal to the equalizer for use in coefficient updating if the equalizer is in a reference-directed mode; and

coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as an error convergence threshold.

20. A method as recited in claim 7, wherein the step of generating the error signal average comprises the steps of:

scaling the error signal; and

generating the error signal average using the scaled error signal.

21. A computer program product for controlling bias in an equalizer, comprising:

a computer readable storage medium having computer readable program code embodied therein, the computer readable program code comprising;

computer readable program code for generating an error signal that corresponds to a difference between a reference signal and an equalizer output signal;

computer readable program code for generating an error signal average; and

computer readable program code for coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as the absolute value of the error signal.

22. A computer program product as recited in claim 21, further comprising:

computer readable program code for generating an average of selected equalizer output signal samples; and

computer readable program code for updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

23. A computer program product as recited in claim 22, wherein the computer readable program code for updating comprises:

computer readable program code for updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average and the equalizer is in a decision-directed mode.

24. A computer program product as recited in claim 22, wherein the computer readable program code for updating comprises:

computer readable program code for updating the reference signal to correspond to the average of selected equalizer output signal samples if the error signal average is less than an error convergence threshold.

25. A computer program product as recited in claim 22, further comprising:

computer readable program code for storing a magnitude of the reference signal for each of a plurality of data frame intervals.

26. A computer program product as recited in claim 25, wherein the computer readable program code for updating comprises:

computer readable program code for updating the magnitude of the reference signal for one of the plurality of data frame intervals to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

27. A computer program product as recited in claim 25, wherein the computer readable program code for generating the average of selected equalizer output signal samples comprises:

computer readable program code for computing a first component by multiplying the magnitude of the reference signal for one of the plurality of data frame intervals by a first weight factor (WF1);

computer readable program code for computing a second component by multiplying the equalizer output signal by a second weight factor (WF2); and

computer readable program code for adding the first component to the second component to generate the average of selected equalizer output signal samples.

28. A computer program product as recited in claim 27, wherein:

the WF1 is between zero and one;

the WF2 is between zero and one; and

the WF1 plus the WF2 is about one.

29. A computer program product as recited in claim 27, wherein the WF1 is about {fraction (63/64)} and the WF2 is about {fraction (1/64)}.

30. A computer program product as recited in claim 21, wherein the computer readable program code for generating the error signal average comprises:

computer readable program code for computing a first error signal average component by multiplying a previous error signal average by a first weight factor (WF1);

computer readable program code for computing a second error signal average component by multiplying the absolute value of the error signal by a second weight factor (WF2); and

computer readable program code for adding the first error signal average component to the second error signal average component to generate the error signal average.

31. A computer program product as recited in claim 30, wherein:

the WF1 is between zero and one;

the WF2 is between zero and one; and

the WF1 plus the WF2 is about one.

32. A computer program product as recited in claim 30, wherein the WF1 is about {fraction (31/32)} and the WF2 is about {fraction (1/32)}.

33. A computer program product as recited in claim 21, wherein the computer readable program code for coupling comprises:

computer readable program code for coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as the absolute value of the error signal;

computer readable program code for coupling the error signal to the equalizer for use in coefficient updating if the equalizer is in a reference-directed mode; and

computer readable program code for coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as an error convergence threshold.

34. A computer program product as recited in claim 21, wherein the computer readable program code for generating the error signal average comprises:

computer readable program code for scaling the error signal; and

computer readable program code for generating the error signal average using the scaled error signal.

35. A bias control system for an equalizer, comprising:

means for generating an error signal that corresponds to a difference between a reference signal and an equalizer output signal;

means for generating an error signal average; and

means for coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as an absolute value of the error signal.

36. A system as recited in claim 35, further comprising:

means for generating an average of selected equalizer output signal samples; and

means for updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

37. A system as recited in claim 36, wherein the means for updating comprises:

means for updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average and the equalizer is in a decision-directed mode.

38. A system as recited in claim 36, wherein the means for updating comprises:

means for updating the reference signal to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average and the error signal average is less than an error convergence threshold.

39. A system as recited in claim 36, further comprising:

means for storing a magnitude of the reference signal for each of a plurality of data frame intervals.

40. A system as recited in claim 39, wherein the means for updating comprises:

means for updating the magnitude of the reference signal for one of the plurality of data frame intervals to correspond to the average of selected equalizer output signal samples if the absolute value of the error signal is greater than the error signal average.

41. A system as recited in claim 39, wherein the means for generating the average of selected equalizer output signal samples comprises:

means for computing a first component by multiplying the magnitude of the reference signal for one of the plurality of data frame intervals by a first weight factor (WF1);

means for computing a second component by multiplying the equalizer output signal by a second weight factor (WF2); and

means for adding the first component to the second component to generate the average of selected equalizer output signal samples.

42. A system as recited in claim 39, wherein:

the WF1 is between zero and one;

the WF2 is between zero and one; and

the WF1 plus the WF2 is about one.

43. A system as recited in claim 39, wherein the WF1 is about {fraction (63/64)} and the WF2 is about {fraction (1/64)}.

44. A system as recited in claim 35, wherein the means for generating the error signal average comprises:

means for computing a first error signal average component by multiplying a previous error signal average by a first weight factor (WF1);

means for computing a second error signal average component by multiplying the absolute value of the error signal by a second weight factor (WF2); and

means for adding the first error signal average component to the second error signal average component to generate the error signal average.

45. A system as recited in claim 44, wherein:

the WF1 is between zero and one;

the WF2 is between zero and one; and

the WF1 plus the WF2 is about one.

46. A system as recited in claim 44, wherein the WF1 is about {fraction (31/32)} and the WF2 is about {fraction (1/32)}.

47. A system as recited in claim 35, wherein the means for coupling comprises:

means for coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as the absolute value of the error signal;

means for coupling the error signal to the equalizer for use in coefficient updating if the equalizer is in a reference-directed mode; and

means for coupling the error signal to the equalizer for use in coefficient updating if the error signal average is at least as great as an error convergence threshold.

48. A system as recited in claim 35, wherein the means for generating the error signal average comprises:

means for scaling the error signal; and

means for generating the error signal average using the scaled error signal.