Crest factor reduction for use in a multiband transmitter

• US 7,313,373 B1
• Filed: 04/21/2005
• Issued: 12/25/2007
• Est. Priority Date: 04/21/2005
• Status: Active Grant

• Current Assignee: Xilinx, Inc. (Advanced Micro Devices, Inc.)
• Original Assignee: Xilinx, Inc. (Advanced Micro Devices, Inc.)
• Inventors: Dick, Christopher H., Tarn, Hai-Jo, Laskharian, Navid
• Primary Examiner(s): NGUYEN, THUAN T
• Application Number:
  US11/111,600
• Time in Patent Office: 978 Days
• Field of Search: 455/552.1, 455/553.1, 455/84, 455/260, 455/208, 455/73, 455/180.1, 455/255, 455/141, 455/127.1
• US Class Current: 455/127.1
• CPC Class Codes: H04B 1/04 Circuits

Abstract

Crest factor reduction in a multiband transmitter is described. Component signals (x_i[n]) are respectively obtained from constituent signals. The component signals (x_i[n]) are respectively associated with sub-bands. A superposed signal associated with the component signals (x_i[n]) is clipped to obtain a clipping noise error signal. The clipping noise error signal is applied to the component signals (x_i[n]) using a least squares estimation to project clipping noise error onto the sub-bands.

Claims

1. A method for crest factor reduction in a multiband transmitter, comprising:

obtaining baseband signals;

combining the baseband signals to provide a superposed signal;

clipping the superposed signal to limit magnitude of the superposed signal;

obtaining a clipping error caused by the clipping;

least squares projecting of the clipping error into sub-band components of the baseband signals to provide projected error components;

filtering the projected error components;

upconverting the projected error components to frequency bands associated with the baseband signals; and

adding the projected error components filtered and upconverted to the superposed signal.

2. The method, according to claim 1, wherein the frequency bands are associated with a channel bandwidth.

3. The method, according to claim 2, wherein the baseband signals are obtained by modulating constituent signals.

4. The method, according to claim 3, wherein the baseband signals occupy non-overlapping frequency bands.

5. The method, according to claim 1, wherein the least squares projecting is an approximation of a least squares fit.

6. The method, according to claim 5, wherein the least squares projecting includes:

multiplying magnitude of the clipping error by cosine of each angle between the clipping error and the baseband signals to obtain clipping noise error magnitudes; and

respectively multiplying the clipping noise error magnitudes by respective phase information for each of the baseband signals.

7. The method, according to claim 6, wherein the least squares projecting includes converting the phase information of the baseband signals from Cartesian coordinate form to polar coordinate form.

8. The method, according to claim 7, wherein the projected error components obtained from respective multiplication of the clipping noise error magnitudes by the phase information for each of the baseband signals is converted from the polar coordinate form to the Cartesian coordinate form.

9. A method for crest factor reduction in a multiband transmitter, comprising:

obtaining a clipped version of a superposed signal;

obtaining an error component associated with the clipped version of the superposed signal;

projecting the error component into sub-bands using a least squares estimation to provide sub-band errors;

spectrally shaping the sub-band errors to comply with an adjacent channel leakage ratio of the multiband transmitter;

frequency modulating resultant noise from the shaping of the sub-band errors to corresponding sub-band frequencies to provide error components; and

combining the error components spectrally shaped and frequency modulated with the superposed signal.

10. The method, according to claim 9, wherein the spectrally shaping includes bandpass filtering.

11. The method, according to claim 9, wherein the spectrally shaping includes:

shifting at least a portion of the sub-band errors to a frequency band; and

bandpass filtering the sub-band errors respectively responsive to the frequency band.

12. The method, according to claim 11, wherein the spectrally shaping includes shifting at least the portion of the sub-band errors back to the sub-bands respectively associated therewith.

13. The method, according to claim 12, wherein the frequency band is associated with a baseband.

14. The method, according to claim 13, wherein the bandpass filtering includes frequency tuning during operation and spectral allocation for both positive and negative frequencies.

15. The method, according to claim 13, wherein the bandpass filtering includes sharing at least one bandpass filter for the spectrally shaping among at least the portion of the sub-band errors.

16. The method, according to claim 9, wherein the combining the error components spectrally shaped and frequency modulated with the superposed signal includes delaying the superposed signal for phase alignment with a combined version of the error components spectrally shaped and frequency modulated.

17. An integrated circuit for crest factor reduction in a multiband transmitter, comprising:

modulators for receiving constituent signals and frequency shifted signals, the constituent signals associated with a signal space, the modulators configured to provide component signals respectively associated with the constituent signals, the component signals occupying respective signal subspaces of the signal space;

a first combiner coupled to receive the component signals and configured to combine the component signals to provide a superposed signal;

a clipping circuit coupled to receive the superposed signal and configured to provide a clipped superposed signal;

a second combiner coupled to receive and to combine the superposed signal with the clipped superposed signal to generate a clipping error signal;

a least squares fit circuit coupled to receive the component signals and the clipping error signal, the least squares fit circuit configured to project clipping noise error associated with the clipping error signal onto the signal subspaces of the component signals;

a bandpass filtering circuit coupled to receive the component signals with the clipping noise error projected thereon and configured to provide noise-shaped subspace error signals respectively responsive to bandpass filtered versions of the component signals to remove at least a portion of the clipping noise error respectively projected thereon;

a third combiner coupled to receive the noise-shaped subspace error signals and configured to combine the noise-shaped subspace error signals to provide a noise-shaped space error signal; and

a fourth combiner coupled to receive the superposed signal and the noise-shaped space error signal at least approximately phase aligned and configured to combine the superposed signal and the noise-shaped error signal to provide a crest factor adjusted version of the superposed signal.

18. The integrated circuit, according to claim 17 wherein the clipping error signal is constrained by the adjacent channel leakage ratio (“

ALCR”

), peak-to-average power ratio (“

PAPR”

), and error vector magnitude (“

EVM”

) parameters.

19. The integrated circuit, according to claim 17, further comprising a coordinate converter coupled to receive the component signals with the clipping noise error projected thereon, the component signals being divided into respective sets of magnitude and phase signals by the least squares fit circuit, the coordinate converter configured to convert the sets of magnitude and phase signals from polar coordinates to Cartesian coordinates.

20. The integrated circuit, according to claim 19, wherein at least the least square fit circuit of the integrated circuit is instantiated in programmably configurable logic.