Method and apparatus for predistortion training in an amplifier utilizing predistortion

• US 7,251,464 B2
• Filed: 10/04/2004
• Issued: 07/31/2007
• Est. Priority Date: 10/16/2003
• Status: Expired due to Fees

• Current Assignee: Motorola Solutions, Inc.
• Original Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Inventors: Thron, Christopher P., Williams, Curtis M., Khan, Andrew M., Opas, George F.
• Primary Examiner(s): VO, NGUYEN THANH
• Application Number:
  US10/958,091
• Publication Number:
  US 20050085199A1
• Time in Patent Office: 1,030 Days
• Field of Search: 455/114.2, 455/114.3, 455/126, 455/127.2, 375/296, 375/297, 330/149
• US Class Current: 455/114.3
• CPC Class Codes: H03F 1/3247 using feedback acting on predistortion circuits H03F1/3264 takes precedence
• CPC Class Codes: H03F 2201/3233 Adaptive predistortion using lookup table, e.g. memory, RAM, ROM, LUT, to generate the predistortion

Abstract

A method and apparatus for predistortion training in an amplifier using predistortion is provided herein. Predistortion takes place by collecting a series of envelope errors and averaging the envelope errors for various amplitude regions. LUT values are modified based on a curve-fit to the average amplitude values for each amplitude region. By utilizing a curve-fitting technique, the pitfalls of modifying individual LUT coefficients is avoided. Particularly, because the errors are collected in relatively broad regions and then averaged, the importance of exact correlation between a measured error and a specific LUT entry is significantly lessened.

Claims

1. A method for predistortion training in an amplifier using predistortion, the method comprising the steps of:

receiving an input signal;

determining an amplitude of the input signal;

associating the amplitude of the input signal with one of a plurality of amplitude regions, wherein the amplitude region is associated with a plurality of LUT indices;

determining an average amplification error for the amplitude region by averaging a plurality of received amplification errors associated with the amplitude region, wherein the plurality of received amplification errors are associated with different LUT indices;

modifying a predistortion LUT based on the average amplification error for the amplitude region; and

predistoring the input signal utilizing the LUT.

2. The method of claim 1 wherein the step of modifying the predistortion LUT based on the average amplification error comprises the step of:

determining if the average amplification error exceeds a predetermined value;

determining an existing curve-fit formula that is utilized by the LUT to generate predistortion values;

modifying the existing curve-fit formula based on the average amplification error; and

updating the LUT with the modified curve-fit formula.

3. The method of claim 2 wherein the step of determining the existing curve-fit formula comprises the step of determining a current set of parametric equations.

4. The method of claim 2 wherein the step of determining the existing curve-fit formula comprises the step of determining a current polynomial equation.

5. The method of claim 1 wherein the step of modifying the predistortion LUT based on the average amplification error comprises the steps of:

determining if the average amplification error exceeds a predetermined value;

determining existing knots in a B-spline curve;

modifying at least one existing knot in the B-spline curve; and

updating the LUT with the modified B-spline curve.

6. The method of claim 1 wherein the step of receiving the input signal comprises the step of receiving modulated baseband data.

7. The method of claim 1 wherein the step of associating the amplitude of the input signal with an amplitude region comprises the step of associating the input signal with region bounded by A_min and A_max such that A_min<

A<

A_max.

8. The method of claim 1 wherein the step of receiving the amplification error for the input signal comprises the step of receiving an amplitude and phase error between the input signal and the amplified signal.

9. An apparatus comprising:

a comparison block determining an amplification error;

a LUT re-trainer receiving an input signal having a given amplitude and amplification error and modifying LUT values based on the given amplitude and the amplification error, wherein the LUT re-trainer modifies LUT values by determining an average error for an amplitude region containing the given amplitude by averaging a plurality of received amplification errors associated with the amplitude region, wherein the plurality of received amplification errors are associated with different LUT indices and modifying a predistortion LUT based on the average error; and

a LUT outputting the modified LUT values to a pre-distorter to be combined with a baseband input signal.

10. The apparatus of claim 9 wherein the LUT re-trainer modifies LUT values based on a curve-fit formula.

11. The apparatus of claim 10 wherein the curve-fit formula comprises curve fitting to a B-spline, and wherein the LUT re-trainer modifies B-spline knots based on the average error.

12. The apparatus of claim 10 wherein the curve-fit formula comprises curve fitting with a set of parametric curves.

13. The apparatus of claim 10 wherein the curve-fit formula comprises curve fitting with a polynomial.

14. The apparatus of claim 9 wherein the comparison block comprises summing circuitry having an amplified signal and the input signal as an input and outputting the amplification error.