System and method for adaptive filter

1. A method for training an adaptive filter, the method comprising:

selecting a filter type from the group consisting of least squares (LS)-GIVEN, least squares QR decomposition (LS-QRD), QR decomposition recursive least squares (QRD), and least mean squares (LMS);

receiving, by a processor from a device, an input signal x and a training reference signal d;

determining, by the processor, a correlation matrix R xd in accordance with the input signal x, the training reference signal d, and the selected filter type;

determining, by the processor, a plurality of coefficients for the adaptive filter in accordance with the correlation matrix R xd ;

adjusting, by the processor, the adaptive filter in accordance with the plurality of coefficients; and

predistorting an input of a power amplifier utilizing the adjusted adaptive filter.

2. The method of claim 1 , further comprising determining a U vector, wherein determining the correlation matrix R xd further comprises determining the correlation matrix R xd in accordance with the U vector.

3. The method of claim 2 , wherein determining the U vector comprises determining the U vector in accordance with the input signal x and the filter type.

4. The method of claim 2 , wherein determining the U vector comprises determining the U vector in accordance with the training reference signal d and the filter type.

5. The method of claim 2 , further comprising determining a size of the U vector.

6. The method of claim 1 , further comprising determining a size of the correlation matrix R xd .

7. The method of claim 1 , further comprising:

determining a first forgetting factor λ 1 in accordance with the filter type;

determining a second forgetting factor λ 2 in accordance with the filter type; and

determining a third forgetting factor λ 3 in accordance with the filter type, wherein determining the correlation matrix R xd further comprises determining the correlation matrix R xd in accordance with the first forgetting factor λ 1 , the second forgetting factor λ 2 , and the third forgetting factor λ 3 .

8. The method of claim 1 , further comprising determining a regularization factor α, wherein determining the correlation matrix R xd further comprises determining the correlation matrix R xd in accordance with the regularization factor α.

9. The method of claim 1 , wherein the correlation matrix R xd consists of a basic block correlation matrix.

10. The method of claim 1 , wherein the correlation matrix R xd comprises:

a first basic block correlation matrix; and

a second basic block correlation matrix.

11. The method of claim 1 , wherein the correlation matrix R xd comprises a partial basic block correlation matrix.

12. The method of claim 1 , wherein the correlation matrix R xd comprises a U vector.

13. A method for training an adaptive filter, the method comprising:

selecting a filter type from the group consisting of least squares (LS)-GIVEN, least squares QR decomposition (LS-QRD), and QR decomposition recursive least squares (QRD);

receiving, by a processor from a device, an input signal x and a training reference signal d;

determining, by the processor, a tri-angle matrix R ud in accordance with the input signal x, the training reference signal d, and the selected filter type;

storing, by the processor, the tri-angle matrix R ud ;

determining, by the processor, a plurality of coefficients for the adaptive filter in accordance with the tri-angle matrix R ud ;

adjusting, by the processor, the adaptive filter in accordance with the plurality of coefficients; and

predistorting an input of a power amplifier utilizing the adjusted adaptive filter.

14. The method of claim 13 , wherein determining the tri-angle matrix R ud comprises:

determining a first plurality of diagonal elements of the tri-angle matrix R ud ; and

determining a second plurality of non-diagonal elements of the tri-angle matrix R ud .

15. The method of claim 14 , wherein determining the second plurality of non-diagonal elements comprises determining a rotation matrix.

16. The method of claim 13 , further comprising determining a size of the tri-angle matrix R ud .

17. The method of claim 13 , further comprising determining a coefficient forgetting factor μ, wherein determining the plurality of coefficients for the adaptive filter comprises determining the plurality of coefficients in accordance with the coefficient forgetting factor μ.

18. The method of claim 13 , wherein the tri-angle matrix R ud consists of a basic rotation matrix.

19. The method of claim 13 , wherein the tri-angle matrix R ud comprises:

a first basic tri-angle matrix R ud j b ; and

a previous tri-angle matrix R ud j−1 .

20. A power amplifier predistortion system comprising:

a processor; and

a computer readable storage medium storing programming for execution by the processor, the programming including instructions to:

selecting a filter type from the group consisting of least squares (LS)-GIVEN, least squares QR decomposition (LS-QRD), QR decomposition recursive least squares (QRD), and least mean squares (LMS);

receive an input signal x and a training reference signal d,

determine a correlation matrix R xd in accordance with the input signal x, the training reference signal d, and the selected filter type,

determine a plurality of coefficients for an adaptive filter in accordance with the correlation matrix R xd ,

adjust the adaptive filter in accordance with the plurality of coefficients; and

predistort an input of a power amplifier utilizing the adjusted adaptive filter.