Carrier recovery system using acquisition and tracking modes and automatic carrier-to-noise estimation

• US 5,471,508 A
• Filed: 08/20/1993
• Issued: 11/28/1995
• Est. Priority Date: 08/20/1993
• Status: Expired due to Fees

• Current Assignee: Hitachi America Limited (Hitachi, Ltd.)
• Original Assignee: Hitachi America Limited (Hitachi, Ltd.)
• Inventors: Koslov, Joshua L.
• Primary Examiner(s): Chin, Stephen
• Assistant Examiner(s): LUTHER, WILLIAM A
• Application Number:
  US08/109,742
• Time in Patent Office: 830 Days
• Field of Search: 375/39, 375/97, 375/119, 375/120, 375/261, 375/344, 375/373, 375/376, 455/226.1, 455/226.3
• US Class Current: 375/344
• CPC Class Codes: H04L 27/3872 Compensation for phase rotation in the demodulated signal

Abstract

A carrier recovery system for use in illustratively a passband QAM (quadrature amplitude modulation) demodulator (10) and specifically such a system, including apparatus and accompanying methods, that employs separate acquisition and tracking modes and automatic carrier-to-noise estimation. In particular, a carrier recovery system (200) is operated in two modes: an acquisition mode to first attain an initial carrier lock, during which simplified reduced constellation slicing is used to determine a quadrant of each incoming symbol; and a tracking (or lock) mode, during which full slicing is used, to accurately track variations in frequency and phase that may occur to a carrier while the system remains locked. The mode changes based upon whether, e.g., the average error energy associated with fully sliced symbols is greater than or less than a pre-defined switching threshold value. The value of the switching threshold changes in a pre-determined manner with changes in noise content, particularly a carrier-to-noise ratio (CNR), of the received symbols. Two appropriately spaced switching threshold values, one for use in each mode, can be used to provide enhanced operational stability. An implementation of a QAM CNR estimator circuit (800) is also provided.

Claims

1. In a communications receiver, apparatus for recovering a carrier signal from a received signal and for producing de-rotated symbols therefrom, said apparatus comprising:

means, responsive to an incoming stream of symbols obtained from said received signal and to a reference carrier signal, for producing a stream of de-rotated symbols;

tracking slicing means for converting each one of said de-rotated symbols into a corresponding first sliced value so as to define a stream of first sliced values, wherein each of said first sliced values defines a corresponding symbol in a pre-selected constellation;

acquisition slicing means for converting ones of said de-rotated symbols into corresponding second sliced values, wherein each one of said second sliced values is specified by a pre-defined attribute of a corresponding one of said de-rotated symbols;

means, operative in response to a mode select signal, for comparing said each one of said de-rotated symbols to a corresponding one of selectively either said first or second sliced values in order to generate a phase error signal;

means, operative in response to said stream of first sliced values and to said stream of de-rotated symbols, for producing a mode error signal;

means, responsive to said mode error signal and a switching threshold value, for producing said mode select signal; and

means, responsive to said phase error signal, for generating said reference carrier.

2. The apparatus in claim 1 further comprising means for varying said switching threshold value in response to variations in either said de-rotated symbols or said incoming stream of symbols.

3. The apparatus in claim 2 wherein each one of said incoming and de-rotated symbols is a quadrature amplitude modulated (QAM) symbol with separate in-phase and quadrature components, each of said first and second sliced values is a complex value, and said reference carrier is a complex signal with separate in-phase and quadrature components.

4. The apparatus in claim 3 wherein said variations are changes in a carrier-to-noise (CNR) of either the stream of de-rotated symbols or said incoming stream of symbols.

5. The apparatus in claim 4 wherein said varying means comprises:

means for estimating the CNR of said stream of de-rotated symbols or said incoming stream of symbols so as to produce a CNR estimate; and

means, responsive to said CNR estimate, for producing said switching threshold value.

6. The apparatus in claim 5 wherein said CNR estimating means comprises means for ascertaining whether said de-rotated symbols, said incoming stream of symbols or a magnitude squared value of either each of said de-rotated symbols or each symbol within said incoming stream of symbols falls within a group of pre-defined ranges and, in response thereto, providing the CNR estimate.

7. The apparatus in claim 6 wherein said switching threshold value producing means comprises a pre-defined look-up table.

8. The apparatus in claim 3 wherein said mode select signal specifies either an acquisition mode or a tracking mode and said mode select signal producing means comprises means for setting said mode select signal to specify said acquisition or tracking mode whenever said mode error signal respectively exceeds or is less than said switching threshold value, and said comparing means further comprises:

means for selecting, in response to said mode select signal, the corresponding one of said first or second sliced values so as to form a selected sliced value; and

phase comparator means for comparing said selected sliced value against said each one de-rotated symbol to generate said phase error signal.

9. The apparatus in claim 8 wherein said sliced signal selecting means is a multiplexer and said phase comparator means comprises one half of a complex multiplier.

10. The apparatus in claim 3 wherein said mode error signal producing means comprises:

means for ascertaining a difference between each one of said de-rotated symbols and each corresponding one of said first sliced values;

means for determining a magnitude squared value of said difference; and

means for averaging said magnitude squared difference value so as to produce an averaged error energy value as said mode error signal.

11. The apparatus in claim 3 wherein said tracking slicing means and said acquisition slicing means are each a two-dimensional quantizer.

12. The apparatus in claim 11 wherein said tracking slicing means separately comprises for the in-phase and quadrature components:

means for quantizing a corresponding component of each of said de-rotated symbols into one of a plurality of pre-defined quantized values; and

means for comparing said one quantized value against pre-defined limit values and producing, in the event the quantized value has a magnitude less than the pre-defined limit values, said quantized value as a corresponding one of said first sliced values or, in the event the magnitude of the quantized value exceeds one of the limit values, the one limit value so exceeded as the corresponding one of said first sliced values.

13. The apparatus in claim 12 further comprising:

means, responsive to said corresponding one de-rotated symbol, for determining whether the corresponding one de-rotated symbol lies beyond a boundary of the pre-selected constellation; and

means, in the event that the one de-rotated symbol lies beyond the boundary and in response to which one of the complex components of the one de-rotated symbol has a larger magnitude than the other, for mapping the complex components of the quantized value for the one corresponding de-rotated symbol into an appropriate QAM symbol situated on the constellation boundary and generating said appropriate QAM symbol, in lieu of said quantized value or said one limit value, as said corresponding one of the first sliced values.

14. The apparatus in claim 11 wherein said acquisition slicing means comprises:

means, responsive to an incoming one of the de-rotated symbols, for squaring magnitudes of corresponding in-phase and quadrature components of said one de-rotated symbol and for adding the magnitudes so squared together to form a resultant sum; and

means for comparing said sum against a pre-defined acquisition slicing threshold value and for generating, in the event the sum exceeds the acquisition slicing threshold value, values for the corresponding in-phase and quadrature components which represent a quadrant in which the de-rotated symbol exists within the pre-selected constellation, and, in the event the sum is less than the acquisition slicing threshold, zero values for the corresponding in-phase and quadrature components.

15. The apparatus in claim 3 wherein said de-rotated symbols producing means is a complex multiplier.

16. The apparatus in claim 3 wherein said pre-selected constellation is one from a plurality of different pre-defined QAM constellations, and wherein said apparatus further comprises means, responsive to a QAM type signal, for changing said pre-selected constellation from any one constellation in said group to any other constellation in said group.

17. The apparatus in claim 16 wherein said pre-defined constellations comprise 16- and 32-QAM.

18. The apparatus in claim 17 wherein said pre-defined constellations further comprise 64-QAM.

19. In a communications receiver, apparatus for recovering a carrier signal from a received signal and for producing de-rotated symbols therefrom, said apparatus comprising:

means, responsive to an incoming stream of symbols obtained from said received signal and to a reference carrier signal, for producing a stream of de-rotated symbols;

tracking slicing means for converting each one of said de-rotated symbols into a corresponding first sliced value so as to define a stream of first sliced values, wherein each of said first sliced values defines a corresponding symbol in a pre-selected constellation;

acquisition slicing means for converting ones of said de-rotated symbols into corresponding second sliced values, wherein each one of said second sliced values is specified by a pre-defined attribute of a corresponding one of said de-rotated symbols;

means, operative in response to a mode select signal, for comparing said each one of said de-rotated symbols to a corresponding one of selectively either said first or second sliced values in order to generate a phase error signal;

means, operative in response to said stream of first sliced values and to said stream of de-rotated symbols, for producing a mode error signal;

means, responsive to said mode error signal and a switching threshold value, for producing said mode select signal;

means for varying said switching threshold value in response to variations in said de-rotated symbols; and

means for generating said reference carrier in response to said phase error signal.

20. The apparatus in claim 19 wherein said variations are changes in a carrier-to-noise (CNR) of either the stream of de-rotated symbols or said incoming stream of symbols.

21. The apparatus in claim 20 wherein said varying means comprises:

means for estimating the CNR of said stream of de-rotated symbols or said incoming stream of symbols so as to produce a CNR estimate; and

means, responsive to said CNR estimate, for producing said switching threshold value.

22. The apparatus in claim 21 wherein each one of said incoming and de-rotated symbols is a quadrature amplitude modulated (QAM) symbol with separate in-phase and quadrature components, each of said first and second sliced values is a complex value, and said reference carrier is a complex signal with separate in-phase and quadrature components.

23. The apparatus in claim 22 wherein said CNR estimating means comprises

means for ascertaining whether said de-rotated symbols, said incoming stream of symbols or a magnitude squared value of either each of said de-rotated symbols or each symbol within said incoming stream of symbols falls within a group of pre-defined ranges and, in response thereto, providing the CNR estimate.

24. The apparatus in claim 23 wherein said switching threshold value producing means comprises a pre-defined look-up table.

25. The apparatus in claim 22 wherein said mode select signal specifies either an acquisition mode or a tracking mode and said mode select signal producing means comprises means for setting said mode select signal to specify said acquisition or tracking mode whenever said mode error signal respectively exceeds or is less than said switching threshold value, and said comparing means further comprises:

means for selecting, in response to said mode select signal, the corresponding one of said first or second sliced values so as to form a selected sliced value; and

phase comparator means for comparing said selected sliced value against said each one de-rotated symbol to generate said phase error signal.

26. The apparatus in claim 22 wherein said mode error signal producing means comprises:

means for ascertaining a difference between each one of said de-rotated symbols and each corresponding one of said first sliced values;

means for determining a magnitude squared value of said difference; and

means for averaging said magnitude squared difference value so as to produce an averaged error energy value as said mode error signal.

27. The apparatus in claim 22 wherein said tracking slicing means and said acquisition slicing means are each a two-dimensional quantizer.

28. The apparatus in claim 27 wherein said tracking slicing means separately comprises for the in-phase and quadrature components:

means for quantizing a corresponding component of each of said de-rotated symbols into one of a plurality of pre-defined quantized values; and

means for comparing said one quantized value against pre-defined limit values and producing, in the event the quantized value has a magnitude less than the pre-defined limit values, said quantized value as a corresponding one of said first sliced values or, in the event the magnitude of the quantized value exceeds one of the limit values, the one limit value so exceeded as the corresponding one of said first sliced values.

29. The apparatus in claim 28 further comprising:

means, responsive to said corresponding one de-rotated symbol, for determining whether the corresponding one de-rotated symbol lies beyond a boundary of the pre-selected constellation; and

means, in the event that the one de-rotated symbol lies beyond the boundary and in response to which one of the complex components of the one de-rotated symbol has a larger magnitude than the other, for mapping the complex components of the quantized value for the one corresponding de-rotated symbol into an appropriate QAM symbol situated on the constellation boundary and generating said appropriate QAM symbol, in lieu of said quantized value or said one limit value, as said corresponding one of the first sliced values.

30. The apparatus in claim 27 wherein said acquisition slicing means comprises:

means, responsive to an incoming one of the de-rotated symbols, for squaring magnitudes of corresponding in-phase and quadrature components of said one de-rotated symbol and for adding the magnitudes so squared together to form a resultant sum; and

means for comparing said sum against a pre-defined acquisition slicing threshold value and for generating, in the event the sum exceeds the acquisition slicing threshold value, values for the corresponding in-phase and quadrature components which represent a quadrant in which the de-rotated symbol exists within the pre-selected constellation, and, in the event the sum is less than the acquisition slicing threshold, zero values for the corresponding in-phase and quadrature components.

31. The apparatus in claim 22 wherein said pre-selected constellation is one from a plurality of different pre-defined QAM constellations, and wherein said apparatus further comprises means, responsive to a QAM type signal, for changing said pre-selected constellation from any one constellation in said group to any other constellation in said group.

32. In a communications receiver, a method for recovering a carrier signal from a received signal and for producing de-rotated symbols therefrom, said method comprising the steps of:

producing, in response to an incoming stream of symbols obtained from said received signal and to a reference carrier signal, a stream of de-rotated symbols;

first converting each one of said de-rotated symbols into a corresponding first sliced value so as to define a stream of first sliced values, wherein each of said first sliced values defines a corresponding symbol in a pre-selected constellation;

second converting ones of said de-rotated symbols into corresponding second sliced values, wherein each one of said second sliced values is specified by a pre-defined attribute of a corresponding one of said de-rotated symbols;

comparing, in response to a mode select signal, said each one of said de-rotated symbols to a corresponding one of selectively either said first or second sliced values in order to generate a phase error signal;

producing, in response to said stream of first sliced values and to said stream of de-rotated symbols, a mode error signal;

producing, in response to said mode error signal and a switching threshold value, said mode select signal; and

generating said reference carrier in response to said phase error signal.

33. The method in claim 32 further comprising the step of varying said switching threshold value in response to variations in either said de-rotated symbols or said incoming stream of symbols.

34. The method in claim 33 wherein each one of said incoming and de-rotated symbols is a quadrature amplitude modulated (QAM) symbol with separate in-phase and quadrature components, each of said first and second sliced values is a complex value, and said reference carrier is a complex signal with separate in-phase and quadrature components.

35. The method in claim 34 wherein said variations are changes in a carrier-to-noise (CNR) of either the stream of de-rotated symbols or said incoming stream of symbols.

36. The method in claim 35 wherein said varying step comprises the steps of:

estimating the CNR of said stream of de-rotated symbols or said incoming stream if symbols so as to produce a CNR estimate; and

producing, in response to said CNR estimate, said switching threshold value.

37. The method in claim 36 wherein said CNR estimating step comprises the step of ascertaining whether said de-rotated symbols, said incoming stream of symbols or a magnitude squared value of either each of said de-rotated symbols or each symbol within said incoming stream of symbols falls within a group of pre-defined ranges and, in response thereto, providing the CNR estimate.

38. The method in claim 34 wherein said mode select signal specifies either an acquisition mode or a tracking mode and said mode select signal producing step comprises the step of setting said mode select signal to specify said acquisition or tracking mode whenever said mode error signal respectively exceeds or is less than said switching threshold value, and said comparing step further comprises the steps of:

selecting, in response to said mode select signal, the corresponding one of said first or second sliced values so as to form a selected sliced value; and

comparing said selected sliced value against said each one de-rotated symbol to generate said phase error signal.

39. The method in claim 34 wherein said mode error signal producing step comprises the steps of:

ascertaining a difference between each one of said de-rotated symbols and each corresponding one of said first sliced values;

determining a magnitude squared value of said difference; and

averaging said magnitude squared difference value so as to produce an averaged error energy value as said mode error signal.

40. The method in claim 34 wherein said first converting step separately comprises, for the in-phase and quadrature components, the steps of:

quantizing a corresponding component of each of said de-rotated symbols into one of a plurality of pre-defined quantized values; and

comparing said one quantized value against pre-defined limit values and producing, in the event the quantized value has a magnitude less than the pre-defined limit values, said quantized value as a corresponding one of said first sliced values or, in the event the magnitude of the quantized value exceeds one of the limit values, the one limit value so exceeded as the corresponding one of said first sliced values.

41. The method in claim 40 further comprising the steps of:

determining, in response to said corresponding one de-rotated symbol, whether the corresponding one de-rotated symbol lies beyond a boundary of the pre-selected constellation; and

in the event that the one de-rotated symbol lies beyond the boundary and in response to which one of the complex components of the one de-rotated symbol has a larger magnitude than the other, mapping the complex components of the quantized value for the one corresponding de-rotated symbol into an appropriate QAM symbol situated on the constellation boundary and generating said appropriate QAM symbol, in lieu of said quantized value or said one limit value, as said corresponding one of the first sliced values.

42. The method in claim 39 wherein said second converting step comprises the steps of:

squaring, in response to an incoming one of the de-rotated symbols, magnitudes of corresponding in-phase and quadrature components of said one de-rotated symbol and adding the magnitudes so squared together to form a resultant sum; and

comparing said sum against a pre-defined acquisition slicing threshold value and generating, in the event the sum exceeds the acquisition slicing threshold value, values for the corresponding in-phase and quadrature components which represent a quadrant in which the de-rotated symbol exists within the pre-selected constellation, and, in the event the sum is less than the acquisition slicing threshold, zero values for the corresponding in-phase and quadrature components.

43. The method in claim 34 wherein said pre-selected constellation is one from a plurality of different pre-defined QAM constellations, and wherein said method further comprises the step of changing, in response to a QAM type signal, said pre-selected constellation from any one constellation in said group to any other constellation in said group.

44. In a communications receiver, a method for recovering a carrier signal from a received signal and for producing de-rotated symbols therefrom, said method comprising:

producing, in response to an incoming stream of symbols obtained from said received signal and to a reference carrier signal, a stream of de-rotated symbols;

first converting each one of said de-rotated symbols into a corresponding first sliced value so as to define a stream of first sliced values, wherein each of said first sliced values defines a corresponding symbol in a pre-selected constellation;

second converting ones of said de-rotated symbols into corresponding second sliced values, wherein each one of said second sliced values is specified by a pre-defined attribute of a corresponding one of said de-rotated symbols;

comparing, in response to a mode select signal, said each one of said de-rotated symbols to a corresponding one of selectively either said first or second sliced values in order to generate a phase error signal;

producing, in response to said stream of first sliced values and to said stream of de-rotated symbols, a mode error signal;

producing, in response to said mode error signal and a switching threshold value, said mode select signal;

varying said switching threshold value in response to variations in said de-rotated symbols; and

generating said reference carrier in response to said phase error signal.

45. The method in claim 44 wherein said variations are changes in a carrier-to-noise (CNR) of either the stream of de-rotated symbols or said incoming stream of symbols.

46. The method in claim 45 wherein said varying step comprises the steps of:

estimating the CNR of said stream of de-rotated symbols or said incoming stream of symbols so as to produce a CNR estimate; and

producing, in response to said CNR estimate, said switching threshold value.

47. The method in claim 46 wherein each one of said incoming and de-rotated symbols is a quadrature amplitude modulated (QAM) symbol with separate in-phase and quadrature components, each of said first and second sliced values is a complex value, and said reference carrier is a complex signal with separate in-phase and quadrature components.

48. The method in claim 47 wherein said CNR estimating step comprises the step of ascertaining whether said de-rotated symbols or said incoming stream of symbols or a magnitude squared value of either each of said de-rotated symbols or each symbol within said incoming stream of symbols falls within a group of pre-defined ranges and, in response thereto, providing the CNR estimate.

49. The method in claim 47 wherein said mode select signal specifies either an acquisition mode or a tracking mode and said mode select signal producing step comprises the step of setting said mode select signal to specify said acquisition or tracking mode whenever said mode error signal respectively exceeds or is less than said switching threshold value, and said comparing step further comprises the steps of:

selecting, in response to said mode select signal, the corresponding one of said first or second sliced values so as to form a selected sliced value; and

comparing said selected sliced value against said each one de-rotated symbol to generate said phase error signal.

50. The method in claim 47 wherein said mode error signal producing step comprises the steps of:

ascertaining a difference between each one of said de-rotated symbols and each corresponding one of said first sliced values;

determining a magnitude squared value of said difference; and

averaging said magnitude squared difference value so as to produce an averaged error energy value as said mode error signal.

51. The method in claim 47 wherein said first converting step separately comprises, for the in-phase and quadrature components, the steps of:

quantizing a corresponding component of each of said de-rotated symbols into one of a plurality of pre-defined quantized values; and

comparing said one quantized value against pre-defined limit values and producing, in the event the quantized value has a magnitude less than the pre-defined limit values, said quantized value as a corresponding one of said first sliced values or, in the event the magnitude of the quantized value exceeds one of the limit values, the one limit value so exceeded as the corresponding one of said first sliced values.

52. The method in claim 51 further comprising the steps of:

determining, in response to said corresponding one de-rotated symbol, whether the corresponding one de-rotated symbol lies beyond a boundary of the pre-selected constellation; and

in the event that the one de-rotated symbol lies beyond the boundary and in response to which one of the complex components of the one de-rotated symbol has a larger magnitude than the other, mapping the complex components of the quantized value for the one corresponding de-rotated symbol into an appropriate QAM symbol situated on the constellation boundary and generating said appropriate QAM symbol, in lieu of said quantized value or said one limit value, as said corresponding one of the first sliced values.

53. The method in claim 47 wherein said second converting step comprises the steps of:

squaring, in response to an incoming one of the de-rotated symbols, magnitudes of corresponding in-phase and quadrature components of said one de-rotated symbol and adding the magnitudes so squared together to form a resultant sum; and

comparing said sum against a pre-defined acquisition slicing threshold value and generating, in the event the sum exceeds the acquisition slicing threshold value, values for the corresponding in-phase and quadrature components which represent a quadrant in which the de-rotated symbol exists within the pre-selected constellation, and, in the event the sum is less than the acquisition slicing threshold, zero values for the corresponding in-phase and quadrature components.

54. The method in claim 47 wherein said pre-selected constellation is one from a plurality of different pre-defined QAM constellations, and wherein said method further comprises the step of changing, in response to a QAM type signal, said pre-selected constellation from any one constellation in said group to any other constellation in said group.

55. Apparatus for estimating a carrier-to-noise ratio (CNR) of symbols recovered from a quadrature amplitude modulated (QAM) signal comprising:

means for ascertaining whether each of said symbols has a symbol magnitude that falls within any one of a group of pre-defined symbol magnitude ranges and, in response thereto, producing an output value indicative of a number of said symbols each of which has a symbol magnitude falling within any one of said ranges, said ranges being defined by corresponding symbol magnitude rings of a QAM constellation containing said symbols, said ascertaining means comprising;

means for generating a magnitude squared value of each of said symbols; and

a multi-level histogram comparator for classifying whether the magnitude squared value of each of said symbols falls within any one of said pre-defined ranges and for producing the output value;

means for filtering the output value so as to generate a filtered value, said filtering means being a infinite impulse response filter, a finite impulse response filter or a periodically reset accumulator; and

a pre-defined look-up table for mapping said filtered value into a CNR estimate.

56. The apparatus in claim 55 wherein for 16-QAM the ranges are, in integers:

0 to 4, and 8 to 12;

or 0 to 8, and 16 to 24.

57. The apparatus in claim 55 wherein for 32-QAM, the ranges are, in integers:

0 to 4, and 8 to 12.

58. A method of estimating a carrier-to-noise ratio (CNR) of symbols recovered from a quadrature amplitude modulated (QAM) signal comprising the steps of:

ascertaining whether each of said symbols has a symbol magnitude that falls within any one of a group of pre-defined symbol magnitude ranges and, in response thereto, producing an output value indicative of a number of said symbols each of which has a symbol magnitude falling within any one of said ranges, said ranges being defined by corresponding symbol magnitude rings of a QAM constellation containing said symbols, said ascertaining step comprising the steps of;

generating a magnitude squared value of each of said symbols; and

classifying, through a multi-level histogram comparator, whether the magnitude squared value of each of said symbols falls within any one of said pre-defined ranges and for producing the output value;

filtering the output value, through an infinite impulse response filter, a finite impulse response filter or a periodically reset accumulator, so as to generate a filtered value; and

mapping, through a pre-defined look-up table, the filtered value into a CNR estimate.

59. The method in claim 58 wherein for 16-QAM the ranges are, in integers:

0 to 4, and 8 to 12;

or 0 to 8, and 16 to 24.

60. The method in claim 58 wherein for 32-QAM, the ranges are, in integers:

0 to 4, and 8 to 12.