Adaptive equalizer with enhanced error quantization

• US 6,418,164 B1
• Filed: 01/14/1999
• Issued: 07/09/2002
• Est. Priority Date: 01/14/1999
• Status: Expired due to Term

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: NxtWave Communications, Inc. (Advanced Micro Devices, Inc.)
• Inventors: Endres, Thomas J, Strolle, Christopher H, Schaffer, Troy A, Hulyalkar, Samir N
• Primary Examiner(s): Pham, Chi
• Assistant Examiner(s): TRAN, KHANH C
• Application Number:
  US09/231,888
• Time in Patent Office: 1,272 Days
• Field of Search: 375/231-233, 375/235, 375/243, 375/245, 375/254, 375/261, 375/270, 375/277, 375/278, 375/285, 375/321, 375/326, 375/335, 375/350, 375/346, 375/240.02, 375/240.03, 708/316, 708/322, 708/323, 455/203, 455/204, 455/109, 348/470, 348/607, 348/729, 348/726, 348/735
• US Class Current: 375/232
• CPC Class Codes: H04L 2025/0342 QAM
• CPC Class Codes: H04L 2025/03624 Zero-forcing
• CPC Class Codes: H04L 27/01 Equalisers baseband equalizers at the transmitter end H04L25/03343; in analogue transmission systems H04B3/04, H04B7/005

Abstract

An adaptive equalizer for use in blind equalization systems to compensate for transmission channel distortion and noise in a digital communication system uses multiple quantization levels for implementation of the Constant Modulus Algorithm (CMA). Different quantization levels are used in different regions of the CMA error function for both passband and baseband equalizers. In one embodiment, a quantizer with a step rise having logarithmic scale is used to digitize the CMA error function. In particular, a quantizer with a step rise in which each level of the quantizer step rise is a power of 2 is used to digitize the CMA error function. In another embodiment, a quantizer with a step rise in which each level of the quantizer step rise is the sum of two or more logarithmic scales is used to digitize the CMA error function. In particular, a quantizer with a step rise wherein each level of the quantizer step rise is the sum of two or more numbers each of which is a power of 2 is used to digitize the CMA error function. The use of logarithmic scales, the sum of logarithmic scales, powers of 2, or the sum of powers of 2 for the quantizer step rise size to digitize the CMA error function make efficient use of shift and add operations to achieve multiplication in the adaptation process.

Claims

1. In a digital communications receiver including an equalization filter having a plurality of adjustable parameters, and an error term calculator for updating said plurality of adjustable parameters, said error term calculator responsive to a received digital signal and the error in said received digital signal, said error being proportional to an error function, a quantization method comprising:

determining first and second regions of said error function;

quantizing said first region of said error function using a first step rise size; and

quantizing said second region of said error function using a second step rise size.

2. A method in accordance with claim 1, further including:

selecting said error function responsive to said received digital signal; and

determining the roots of said error function, wherein said roots define the boundaries of said first and second regions of said error function.

3. A method in accordance with claim 1, wherein said first region of said error function using said first step rise size is quantized using a logarithmic scale.

4. A method in accordance with claim 3, wherein said second region of said error function using said second step rise size is quantized using a logarithmic scale.

5. A method in accordance with claim 1, wherein said first step rise size is selected to be a power of 2.

6. A method in accordance with claim 5, wherein said second step rise size is selected to be a power of 2.

7. A method in accordance with claim 1, wherein said first step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

8. A method in accordance with claim 7, wherein said second step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

9. A method in accordance with claim 1, wherein said first step rise size is equal to 40.

10. A method in accordance with claim 1, wherein said second step rise size is equal to 128.

11. A method in accordance with claim 1, wherein the boundary between said first and second regions of said error function is the square root of Godard'"'"'s radius.

36. A method in accordance with claim 5, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said method further comprising performing said multiply operation by only one shift operation.

38. A method in accordance with claim 7, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said method further comprising performing said multiply operation by only two shift operations and one add operation.

12. In a digital communications receiver including an equalization filter having a plurality of adjustable parameters, and an error term calculator for updating said plurality of adjustable parameters, said error term calculator responsive to a received digital signal and the error in said received digital signal, said error being proportional to an error function, a quantization apparatus comprising:

means for determining first and second regions of said error function;

means for quantizing said first region of said error function using a first step rise size; and

means for quantizing said second region of said error function using a second step rise size.

13. An apparatus in accordance with claim 12, further including:

means for selecting said error function responsive to said received digital signal; and

means for determining the roots of said error function, wherein said roots define the boundaries of said first and second regions of said error function.

14. An apparatus in accordance with claim 12, wherein said first region of said error function using said first step rise size is quantized using a logarithmic scale.

15. An apparatus in accordance with claim 14, wherein said second region of said error function using said second step rise size is quantized using a logarithmic scale.

16. An apparatus in accordance with claim 12, wherein said first step rise size is selected to be a power of 2.

17. An apparatus in accordance with claim 16, wherein said second step rise size is selected to be a power of 2.

18. An apparatus in accordance with claim 12, wherein said first step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

19. An apparatus in accordance with claim 18, wherein said second step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

20. An apparatus in accordance with claim 12, wherein said first step rise size is equal to 40.

21. An apparatus in accordance with claim 12, wherein said second

step rise size is equal to 128.

22. An apparatus in accordance with claim 13, wherein the boundary between said first and second regions of said error function is the square root of Godard'"'"'s radius.

40. An apparatus in accordance with claim 16, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said apparatus further comprising means for performing said multiply operation by only one shift operation.

42. An apparatus in accordance with claim 18, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said apparatus further comprising means for performing said multiply operation by only two shift operations and one add operation.

23. In a digital communications receiver including an equalization filter having a plurality of adjustable parameters, and an error term calculator for updating said plurality of adjustable parameters, said error term calculator responsive to a received digital signal and the error in said received digital signal, said error being proportional to an error function, a quantization method comprising:

quantizing said error function using a step rise size according to a logarithmic scale to provide a quantized error function; and

updating said plurality of adjustable parameters using said quantized error function.

24. A method in accordance with claim 23, wherein said step rise size is selected to be a power of 2.

25. A method in accordance with claim 23, wherein said step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

37. A method in accordance with claim 24 wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said method further comprising performing said multiply operation by only one shift operation.

39. A method in accordance with claim 25, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said method further comprising performing said multiply operation by only two shift operations and one add operation.

26. In a digital communications receiver including an equalization filter having a plurality of adjustable parameters, and an error term calculator for updating said plurality of adjustable parameters, said error term calculator responsive to a received digital signal and the error in said received digital signal, said error being proportional to an error function, a quantization apparatus comprising:

means for quantizing said error function using a step rise size according to a logarithmic scale to provide a quantized error function; and

means for updating said plurality of adjustable parameters using said quantized error function.

27. An apparatus in accordance with claim 26, wherein said step rise size is selected to be a power of 2.

28. An apparatus in accordance with claim 26, wherein said step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

41. An apparatus in accordance with claim 27, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said apparatus further comprising means for performing said multiply operation by only one shift operation.

43. An apparatus in accordance with claim 28, wherein said error term calculator for updating said plurality of adjustable parameters includes at least one multiply operation, said apparatus further comprising means for performing said multiply operation by only two shift operations and one add operation.

29. An equalizer for a digital communications receiver for receiving digital symbols, said equalizer comprising:

an equalization filter having a plurality of adjustable parameters, and an output terminal providing said received digital symbols;

a slicer having a respective input and output, the input of said slicer being coupled to the output of said equalization filter, said slicer being set for the nominal decision threshold levels corresponding to each of said digital symbols;

an error term calculator for updating said plurality of adjustable parameters of said equalization filter, said error term calculator responsive to said received digital symbols and to the output of said slicer to determine an error function;

said error term calculator further responsive to said received digital symbols to determine first and second regions of said error function; and

a quantizer having a respective input and output, said quantizer further having a first step rise size to digitize said error function in said first region and a second step rise size to digitize said error function in said second region, wherein the output of said quantizer is used in said error term calculator to update said plurality of adjustable parameters of said equalization filter.

30. An equalization filter in accordance with claim 29, wherein at least one of said first and second regions of said error function are respectively quantized using a logarithmic scale.

31. A method in accordance with claim 29, wherein at least one of said first and second step rise size is selected to be a power of 2.

32. A method in accordance with claim 29, wherein at least one of said first and second step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.

33. An equalizer for a digital communications receiver for

receiving digital symbols, said equalizer comprising:

an equalization filter having a plurality of adjustable parameters, and an output terminal providing said received digital symbols;

a slicer having a respective input and output, the input of said slicer being coupled to the output of said equalization filter, said slicer being set for the nominal decision threshold levels corresponding to each of said digital symbols;

an error term calculator for updating said plurality of adjustable parameters of said equalization filter, said error term calculator responsive to said received digital symbols and to the output of said slicer to determine an error function; and

a quantizer having a respective input and output, said quantizer further having a step rise size to digitize said error function, said step rise size being a logarithmic scale, wherein the output of said quantizer is used in said error term calculator to update said plurality of adjustable parameters of said equalization filter.

34. A quantizer in accordance with claim 33, wherein said step rise size is selected to be a power of 2.

35. A method in accordance with claim 34, wherein said step rise size is selected to be the sum of at least two numbers, each of which numbers is a power of 2.