Adaptive digital pre-distortion using amplifier model that incorporates frequency-dependent non-linearities

• US 6,801,086 B1
• Filed: 09/17/2002
• Issued: 10/05/2004
• Est. Priority Date: 04/03/2002
• Status: Expired due to Fees

• Current Assignee: CommScope Technologies, LLC (CommScope Holding Company, Inc.)
• Original Assignee: Andrew Corporation (CommScope Holding Company, Inc.)
• Inventors: Chandrasekaran, Rajiv
• Primary Examiner(s): Nguyen, Khanh Van
• Application Number:
  US10/245,008
• Time in Patent Office: 749 Days
• Field of Search: 330/2, 330/52, 330/149, 330/296, 330/297, 375/296, 375/297
• US Class Current: 330/140
• CPC Class Codes: H03F 1/3247 using feedback acting on predistortion circuits H03F1/3264 takes precedence
• CPC Class Codes: H03F 1/3252 using multiple parallel paths between input and output H03F1/3258, H03F1/3282, H03F1/3294 take precedence
• CPC Class Codes: H03F 1/3258 based on polynomial terms

Abstract

A pre-distorter pre-distorts an input signal prior to being applied to an amplifier in order to reduce spurious emissions in the resulting amplified signal. The pre-distorter implements an inverted version of a model of the amplifier that models both the frequency independent (FI) characteristics of the amplifier as well as the frequency-independent (FD) characteristics of the amplifier. Techniques and architectures are presented for (1) generating and updating the model, (2) inverting the model, and (3) updating the inverted model.

Claims

1. A method for pre-distorting a signal for amplification, the method comprising the steps of:

(a) receiving an input signal; and

(b) applying pre-distortion to the input signal to generate a pre-distorted signal, such that, when the pre-distorted signal is applied to an amplifier to generate an amplified signal, the pre-distortion reduces spurious emissions in the amplified signal, wherein;

the pre-distortion is generated using an inverse of a model of the amplifier; and

the model comprises a model of frequency-independent (FI) characteristics of the amplifier (an FI model) in combination with a model of frequency-dependent (FD) characteristics of the amplifier (an FD model).

2. The invention of claim 1, further comprising the step of generating at least one of the FI model and the FD model.

3. The invention of claim 2, wherein the FD model is generated by estimating a plurality of transfer functions in a sequence corresponding to increasing order of the transfer function.

4. The invention of claim 2, wherein the model is generated by:

(1) estimating the FI model based on an estimate of the frequency-independent gain of the amplifier;

(2) estimating a first-order transfer function for the FD model using the estimated FI model; and

(3) estimating a second-order transfer function for the FD model using the estimated FI model and the estimated first-order transfer function.

5. The invention of claim 4, wherein:

the FI model is estimated by minimizing a cost function that is dependent on the gain of the amplifier;

the first-order transfer function is estimated by minimizing a cost function that is dependent on the estimated FI model; and

the second-order transfer function is estimated by minimizing a cost function that is dependent on the estimated FI model and the estimated first-order transfer function.

6. The invention of claim 4, wherein the model is further generated by estimating one or more transfer functions for the FD model of order greater than two, wherein a transfer function of order n is estimated using the estimated FI model and each estimated transfer function of order less than n.

7. The invention of claim 2, further comprising the step of adaptively updating the model.

8. The invention of claim 7, wherein the model is adaptively updated by adaptively updating a plurality of transfer functions in a sequence corresponding to increasing order of the transfer function.

9. The invention of claim 1, wherein the FD model comprises a plurality of different-order transfer functions, each of which is configured to operate on a different-order distortion product based on the input signal, wherein:

the output from the FI model is applied to a first-order transfer function; and

the outputs from the plurality of transfer functions are summed to generate an estimate of the amplified signal.

10. The invention of claim 9, wherein the FD model comprises a second-order transfer function that is configured to operate on a difference between the input signal and the output from the FI model.

11. The invention of claim 9, wherein the FD model comprises one or more transfer functions having order two or greater, each of which is configured to operate on a distortion product having a corresponding order of two or greater.

12. The invention of claim 11, wherein the FD model comprises two or more transfer functions having order two or greater, each of which is configured to operate on a distortion product having a corresponding order of two or greater.

13. The invention of claim 9, wherein, in generating the model, the plurality of transfer functions are estimated in a sequence corresponding to increasing order.

14. The invention of claim 1, wherein the inverse of the model comprises:

(1) zero, one, or more negated high-order transfer function elements, each corresponding to a negation of one of one or more transfer functions of the FD model of order greater than one and configured to operate on a corresponding distortion product;

(2) an inverted first-order transfer function element corresponding to an inverse of a first-order transfer function of the FD model and configured to operate on the sum of the input signal and the zero, one, or more outputs from the zero, one, or more negated high-order transfer function elements; and

(3) an inverted FI model element corresponding to an inverse of the FI model and configured to operate on the output from the inverted first-order transfer function to generate the pre-distorted signal.

15. The invention of claim 14, wherein:

the zero, one or more negated high-order transfer function elements include a negated second-order transfer function element comprising;

(A) a first component configured to generate a measure of the input signal;

(B) a first multiplier configured to multiply the input signal by the measure of the input signal to generate a second-order distortion product; and

(C) a first filter configured to apply a negated second-order transfer function to the second-order distortion product;

the inverted first-order transfer function element comprises a second filter configured to apply the inverted first-order transfer function to the sum of the input signal and the one or more outputs from the one or more negated high-order transfer function elements; and

the inverted FI model element comprises;

(A) a second component configured to generate a measure of the output from the second filter;

(B) a look-up table configured to retrieve one or more pre-distortion components based on the measure of the output from the second filter; and

(C) a second multiplier configured to multiply the output from the second filter by the one or more pre-distortion components to generate the pre-distorted signal.

16. The invention of claim 15, wherein:

the first component is configured to generate a measure of the amplitude or the power of the input signal; and

the first filter is a finite impulse response (FIR) filter.

17. The invention of claim 16, wherein the second filter is an FIR filter corresponding to the inverse of a first-order transfer function.

18. The invention of claim 16, wherein:

the inverted first-order transfer function further comprises a baseband-to-RF converter preceding the second filter;

the second filter is an RF filter corresponding to the inverse of a first-order transfer function; and

the second multiplier is a vector modulator.

19. The invention of claim 1, wherein:

the inverse of the model comprises an inverse of the FI model;

the inverse of the FI model generates an inverted FI model output having an amplitude and a phase;

the magnitude of the inverted FI model output is a function of the magnitude of the input signal; and

the phase of the inverted FI model output is a function of both the magnitude and phase of the input signal.

20. The invention of claim 19, wherein:

the magnitude of the inverted FI model output is derived using the input signal magnitude as an index into an AM/AM look-up table;

the phase of the inverted FI model output is derived by summing the input signal phase and a value retrieved using the input signal magnitude as an index into an AM/PM look-up table.

21. The invention of claim 20, wherein at least one of the look-up tables is adaptively updated by minimizing an error between the input signal and a feedback signal based on the amplified signal.

22. The invention of claim 21, wherein the at least one look-up table is updated by applying a recursive estimation technique to the look-up table.

23. The invention of claim 22, wherein the recursive estimation technique is based on non-linear least mean squares estimation.

24. The invention of claim 1, wherein the amplifier model is a linear combination of the FI model and the FD model, where the FI model is distinct from the FD model.

25. An apparatus for pre-distorting a signal for amplification, wherein the apparatus is configured to:

(a) receive an input signal; and

(b) apply pre-distortion to the input signal to generate a pre-distorted signal, such that, when the pre-distorted signal is applied to an amplifier to generate an amplified signal, the pre-distortion reduces spurious emissions in the amplified signal, wherein;

the pre-distortion is generated using an inverse of a model of the amplifier, and the model comprises an FI model of frequency-independent characteristics of the amplifier in combination with an FD model of frequency-dependent characteristics of the amplifier.

31. The invention of claim 25, wherein the amplifier model is a linear combination of the FI model and the FD model, where the FI model is distinct from the FD model.

26. An apparatus for pre-distorting a signal for amplification, the apparatus comprising:

(a) One or more high-order transfer function elements, each configured to apply a transfer function of a different order greater than one to a corresponding distortion product for an input signal;

(b) a summation node configured to sum the input signal and the output of each high-order transfer function element;

(c) an inverted transfer function element configured to apply an inverted first-order transfer function to the output of the summation node; and

(d) an inverted FI element configured to apply an inverse of a model of a frequency-independent (FI) gain of the amplifier to the output of the inverted transfer function element to generate a pre-distorted signal, such that, when the pre-distorted signal is applied to an amplifier to generate an amplified signal, spurious emissions in the amplified signal are reduced.

27. The invention of claim 26, wherein:

the one or more high-order transfer function elements include a second-order transfer function element comprising;

(A) a first component configured to generate a measure of the input signal;

(B) a first multiplier configured to multiply the input signal by the measure of the input signal to generate a second-order distortion product; and

(C) a first filter configured to apply a second-order transfer function to the second-order distortion product;

the inverted first-order transfer function element comprises a second filter configured to apply the inverted first-order transfer function to the sum of the input signal and the outputs from the one or more high-order transfer function elements; and

the inverted FI element comprises;

(A) a second component configured to generate a measure of the output from the second filter;

(B) a look-up table configured to retrieve one or more pre-distortion components based on the measure of the output from the second filter; and

(C) a second multiplier configured to multiply the output from the second filter by the one or more pre-distortion components to generate the pre-distorted signal.

28. The invention of claim 27, wherein:

the first component is configured to generate a measure of the amplitude or the power of the input signal; and

the first filter is a finite impulse response (FIR) filter.

29. The invention of claim 28, wherein the second filter is an FIR filter corresponding to the inverse of a first-order transfer function.

30. The invention of claim 28, wherein:

the inverted first-order transfer function further comprises a baseband-to-RF converter preceding the second filter;

the second filter is an RF filter corresponding to the inverse of a first-order transfer function; and

the second multiplier is a vector modulator.

32. A method for generating a model of an amplifier, wherein the model comprises an FI model of frequency-independent characteristics of the amplifier in combination with an FD model of frequency-dependent characteristics of the amplifier, wherein the model is generated by estimating a plurality of transfer functions for the FD model in a sequence corresponding to increasing order of the transfer function.

33. The invention of claim 32, wherein the model is generated by:

(1) estimating the FI model based on an estimate of the frequency-independent gain of the amplifier;

(2) estimating a first-order transfer function for the FD model using the estimated FI model; and

(3) estimating a second-order transfer function for the FD model using the estimated FI model and the estimated first-order transfer function.

34. The invention of claim 33, wherein:

the FI model is estimated by minimizing a cost function that is dependent on the gain of the amplifier;

the first-order transfer function is estimated by minimizing a cost function that is dependent on the estimated FI model; and

the second-order transfer function is estimated by minimizing a cost function that is dependent on the estimated FI model and the estimated first-order transfer function.

35. The invention of claim 33, wherein the model is further generated by estimating one or more transfer functions for the FD model of order greater than two, wherein a transfer function of order n is estimated using the estimated FI model and each estimated transfer function of order less than n.

36. The invention of claim 32, further comprising the step of adaptively updating the model.

37. The invention of claim 36, wherein the model is adaptively updated by adaptively updating a plurality of transfer functions in a sequence corresponding to increasing order of the transfer function.

38. The invention of claim 32, wherein the amplifier model is a linear combination of the FI model and the FD model, where the FI model is distinct from the FD model.

39. A method for adaptively updating a pre-distorter configured to pre-distort an input signal for application to an amplifier configured to generate an amplified signal, the pre-distorter comprising at least one look-up table (LUT), wherein the at least one LUT is adaptively updated by minimizing an error between the input signal and a feedback signal based on the amplified signal, wherein the at least one look-un table is undated by applying a recursive estimation technique to the at least one look-up table.

40. The invention of claim 39, wherein the recursive estimation technique is based on nonlinear least mean squares estimation.