Adaptive predistortion of non-linear amplifiers with burst data
View Patent ↗Apparatus and methods control predistortion of an RF transmitter. A base station or a mobile station can utilize predistortion to improve linearity characteristics of the RF power amplifier. When used effectively, predistortion limits spectral growth such that the amplified signal complies with regulatory requirements. With respect to bursty signals, specific improvement techniques are disclosed. A first technique is related to adaptation using only a smaller subset of samples of a burst. A second technique is related to selective application of digital predistortion, such as, only under high power conditions for a power amplifier. A third technique is directed to adaptation of less than all of the coefficients. These improvements permit the use of a smaller and less expensive amplifier for a given power class and can lengthen battery life for a mobile unit.
1. An apparatus for RF power amplifier predistortion, the apparatus comprising:
a predistorter configured to receive a reference signal, the predistorter configured to predistort the reference signal to generate a predistorted signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network; and
an adaptive estimator configured to receive an observed signal that is derived from an output of an RF power amplifier, wherein the adaptive estimator is also configured to receive the reference signal, wherein the adaptive estimator is configured to update one or more coefficients of the predistorter based on only a smaller subset of samples of the bursts for the reference signal and for the observed signal than all of the samples, wherein the smaller subset corresponds to a region of the burst that is expected to encounter the most distortion when amplified.
2. The apparatus of claim 1 , wherein a burst comprises at least a payload of data.
3. The apparatus of claim 1 , wherein the smaller subset of samples is selected to end before the RF power amplifier has achieved steady state operation.
4. The apparatus of claim 3 , wherein the smaller subset of samples is selected to start at a beginning of a burst.
5. The apparatus of claim 1 , wherein the smaller subset consists of only about a 5-15% of samples of a burst.
6. The apparatus of claim 5 , wherein the smaller subset is selected to start at a beginning of a burst.
7. The apparatus of claim 5 , wherein the burst further comprises an overhead portion, wherein the smaller subset comprises at least a preamble of the burst.
8. An apparatus for RF power amplifier predistortion, the apparatus comprising:
a predistorter configured to receive a reference signal, the predistorter configured to predistort the reference signal to generate a predistorted signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network; and
an adaptive estimator configured to receive an observed signal that is derived from an output of an RF power amplifier, wherein the adaptive estimator is also configured to receive the reference signal, wherein the adaptive estimator is configured to update one or more coefficients of the predistorter based on only a smaller subset of samples of the bursts for the reference signal and for the observed signal than all of the samples, wherein the smaller subset corresponds to a region of the burst requiring the lowest error vector magnitude.
9. An apparatus for RF power amplifier predistortion, the apparatus comprising:
a predistorter configured to receive a reference signal, the predistorter configured to predistort the reference signal to generate a predistorted signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network; and
an adaptive estimator configured to receive an observed signal that is derived from an output of an RF power amplifier, wherein the adaptive estimator is also configured to receive the reference signal, wherein the adaptive estimator is configured to update one or more coefficients of the predistorter based on only a smaller subset of samples of the bursts for the reference signal and for the observed signal than all of the samples, wherein the smaller subset corresponds to a region of the burst that is to be amplified to the highest output power within the burst.
10. A method of reducing distortion in a RF power amplifier, the method comprising:
receiving a reference signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network;
predistorting the reference signal with a set of predistortion coefficients to generate a predistorted signal, wherein a predistortion characteristic is intended to be complementary to a distortion of an RF power amplifier;
receiving an observed signal, wherein the observed signal is derived from an output of the RF power amplifier; and
adaptively updating one or more coefficients of the set of predistortion coefficients based at least partly on a comparison between selected samples of the reference signal and the observed signal, wherein the selected samples consist of only a smaller subset of samples of the bursts for the reference signal and for the observed signal than all of the samples, wherein the smaller subset corresponds to a region of the burst that is expected to encounter the most distortion when amplified.
11. The method of claim 10 , wherein a burst comprises at least a payload of data.
12. The method of claim 10 , wherein the smaller subset of samples is selected to end before the RF power amplifier has achieved steady state operation.
13. The method of claim 12 , wherein the smaller subset of samples is selected to start at a beginning of a burst.
14. The method of claim 10 , wherein the smaller subset consists of only about a 5-15% of samples of a burst of information.
15. The method of claim 14 , wherein the smaller subset is selected to start at a beginning of the burst.
16. The method of claim 14 , wherein the burst further comprises an overhead portion, wherein the smaller subset comprises at least a preamble of the burst.
17. A method of reducing distortion in a RF power amplifier, the method comprising:
receiving a reference signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network;
predistorting the reference signal with a set of predistortion coefficients to generate a predistorted signal, wherein a predistortion characteristic is intended to be complementary to a distortion of an RF power amplifier;
receiving an observed signal, wherein the observed signal is derived from an output of the RF power amplifier; and
adaptively updating one or more coefficients of the set of predistortion coefficients based at least partly on a comparison between selected samples of the reference signal and the observed signal, wherein the selected samples consist of only a smaller subset of samples of the bursts for the reference signal and for the observed signal than all of the samples, wherein the smaller subset corresponds to a region of the burst requiring the lowest error vector magnitude.
18. A method of reducing distortion in a RF power amplifier, the method comprising:
receiving a reference signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network;
predistorting the reference signal with a set of predistortion coefficients to generate a predistorted signal, wherein a predistortion characteristic is intended to be complementary to a distortion of an RF power amplifier;
receiving an observed signal, wherein the observed signal is derived from an output of the RF power amplifier; and
adaptively updating one or more coefficients of the set of predistortion coefficients based at least partly on a comparison between selected samples of the reference signal and the observed signal, wherein the selected samples consist of only a smaller subset of samples of the bursts for the reference signal and for the observed signal than all of the samples, wherein the smaller subset corresponds to a region of the burst that is to be amplified to the highest output power within the burst.
19. An apparatus for RF power amplifier predistortion, the apparatus comprising:
a predistorter configured to receive a reference signal and configured to be able to predistort the reference signal to generate a predistorted signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network; and
a control circuit coupled to the predistorter, the control circuit configured to determine whether predistortion is to be applied to a particular burst based at least partly on a peak RF power level of an RF burst associated with the particular burst, and to apply predistortion to the particular burst only when the peak RF power level exceeds a threshold, and not to apply predistortion to the particular burst when the peak RF power level is below the threshold.
20. The apparatus of claim 19 , wherein the predistortion circuit is configured to switch among at least predistortion or a complete bypass of predistortion based on a state of a control signal from the control circuit.
21. The apparatus of claim 19 , wherein the predistortion circuit is configured to select from among at least predistortion or simple gain scaling based on a state of control signal from the control circuit.
22. The apparatus of claim 19 , wherein the threshold is at least 6 decibels below the maximum transmit power.
23. A method for control of RF power amplifier predistortion, the method comprising:
receiving a reference signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network;
determining whether predistortion is to be applied to a particular burst based at least partly on a peak RF power level of an RF burst associated with the particular burst, and to apply predistortion to the particular burst only when the peak RF power level exceeds a threshold, and not to apply predistortion to the particular burst when the peak RF power level is below the threshold; and
selectively predistorting the particular burst based on the determination, wherein a predistortion characteristic is intended to be complementary to a distortion of an RF power amplifier.
24. The method of claim 23 , wherein selectively predistorting comprises selecting among at least predistorting or completely bypassing of predistortion.
25. The method of claim 23 , wherein selectively predistorting comprises selecting from among at least predistortion or gain scaling.
26. The method of claim 23 , wherein the threshold is at least 6 decibels below the maximum transmit power.
27. An apparatus for RF power amplifier predistortion, the apparatus comprising:
a predistorter configured to receive a reference signal and configured to predistort the reference signal to generate a predistorted signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network; and
an adaptive estimator configured to receive an observed signal that is derived from an output of an RF power amplifier, wherein the adaptive estimator is also configured to receive the reference signal, wherein the adaptive estimator is configured to update coefficients for both a first portion of coefficients and a second portion of coefficients of the predistorter until at least convergence is achieved, and is then configured to adaptively update the first portion of the coefficients and not to update the second portion of coefficients, the first portion comprising at least coefficients corresponding to peak amplitude operation, wherein there is no overlap between the first portion and the second portion.
28. The apparatus of claim 27 , wherein the adaptive estimator is further configured to adaptively update the second portion momentarily upon determining that a change in operating condition has occurred.
29. A method of reducing distortion in a RF power amplifier, the method comprising:
receiving a reference signal, wherein the reference signal contains information in discrete bursts for transmission in a time-division duplex (TDD) network;
predistorting the reference signal with a set of predistortion coefficients to generate a predistorted signal, wherein a predistortion characteristic is intended to be complementary to a distortion of an RF power amplifier;
receiving an observed signal, wherein the observed signal is derived from an output of the RF power amplifier; and
adaptively updating predistortion coefficients based at least partly on a comparison between samples of the reference signal and the observed signal, wherein the predistortion coefficients are divided into at least a first portion and a second portion, wherein both the first portion and the second portion are adaptively updated until initial convergence is achieved, and wherein only the first portion and not the second portion are adaptively adjusted thereafter, wherein the first portion comprises at least coefficients corresponding to peak amplitude operation.
30. The method of claim 29 , wherein the second portion is also adaptively adjusted momentarily.
31. The apparatus of claim 8 , wherein a burst comprises at least a payload of data.
32. The apparatus of claim 8 , wherein the smaller subset of samples is selected to end before the RF power amplifier has achieved steady state operation.
33. The apparatus of claim 32 , wherein the smaller subset of samples is selected to start at a beginning of a burst.
34. The apparatus of claim 8 , wherein the smaller subset consists of only about a 5-15% of samples of a burst.
35. The apparatus of claim 34 , wherein the smaller subset is selected to start at a beginning of a burst.
36. The apparatus of claim 34 , wherein the burst further comprises an overhead portion, wherein the smaller subset comprises at least a preamble of the burst.
37. The apparatus of claim 9 , wherein a burst comprises at least a payload of data.
38. The apparatus of claim 9 , wherein the smaller subset of samples is selected to end before the RF power amplifier has achieved steady state operation.
39. The apparatus of claim 38 , wherein the smaller subset of samples is selected to start at a beginning of a burst.
40. The apparatus of claim 9 , wherein the smaller subset consists of only about a 5-15% of samples of a burst.
41. The apparatus of claim 40 , wherein the smaller subset is selected to start at a beginning of a burst.
42. The apparatus of claim 40 , wherein the burst further comprises an overhead portion, wherein the smaller subset comprises at least a preamble of the burst.
43. The method of claim 17 , wherein a burst comprises at least a payload of data.
44. The method of claim 17 , wherein the smaller subset of samples is selected to end before the RF power amplifier has achieved steady state operation.
45. The method of claim 44 , wherein the smaller subset of samples is selected to start at a beginning of a burst.
46. The method of claim 17 , wherein the smaller subset consists of only about a 5-15% of samples of a burst of information.
47. The method of claim 46 , wherein the smaller subset is selected to start at a beginning of the burst.
48. The method of claim 46 , wherein the burst further comprises an overhead portion, wherein the smaller subset comprises at least a preamble of the burst.
49. The method of claim 18 , wherein a burst comprises at least a payload of data.
50. The method of claim 18 , wherein the smaller subset of samples is selected to end before the RF power amplifier has achieved steady state operation.
51. The method of claim 50 , wherein the smaller subset of samples is selected to start at a beginning of a burst.
52. The method of claim 18 , wherein the smaller subset consists of only about a 5-15% of samples of a burst of information.
53. The method of claim 52 , wherein the smaller subset is selected to start at a beginning of the burst.
54. The method of claim 52 , wherein the burst further comprises an overhead portion, wherein the smaller subset comprises at least a preamble of the burst.