Processing signals

• US 9,042,573 B2
• Filed: 11/30/2011
• Issued: 05/26/2015
• Est. Priority Date: 09/30/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Skype S.Ã R.L. (Microsoft Corporation)
• Inventors: hgren, Per, Sorensen, Karsten Vandborg
• Primary Examiner(s): Chin, Vivian
• Assistant Examiner(s): FAHNERT, FRIEDRICH
• Application Number:
  US13/308,165
• Publication Number:
  US 20130083942A1
• Time in Patent Office: 1,273 Days
• Field of Search: 381/92, 381/93, 381/95, 381/96, 381/94.1, 381/94.3, 381/94.5, 700/94
• US Class Current: 381/92
• CPC Class Codes: G01S 3/8006 Multi-channel systems specially adapted for direction-finding, i.e. having a single aerial system capable of giving simultaneous indications of the directions of different signals
• CPC Class Codes: G01S 3/8083 determining direction of source
• CPC Class Codes: G10L 2021/02166 Microphone arrays; Beamforming
• CPC Class Codes: H04R 2499/11 Transducers incorporated or for use in hand-held devices, e.g. mobile phones, PDA's, camera's
• CPC Class Codes: H04R 3/005 for combining the signals of two or more microphones specially adapted for hearing aids H04R25/407

Abstract

Beamformer coefficients may include a plurality of sets of theoretical statistical data for theoretical signals. Each theoretical signal may have its own particular attributes. The statistical data may be used in computing beamformer coefficients for application by a beamformer to signals received at a device. Signals are received at an input of the device. A respective plurality of weights is determined, for the theoretical statistical data sets, based on an analysis of the extent to which the signals have the particular attributes of the theoretical signals. The theoretical are retrieved, and a statistical data set is calculated for the signals by performing a weighted sum of the theoretical statistical data sets using the determined respective plurality of weights. Beamformer coefficients are computed based on the calculated statistical data set for the signals, which are used by a beamformer to the signals for generating a beamformer output.

Claims

1. A method of computing beamformer coefficients at a device, the method comprising:

storing a plurality of theoretical covariance matrices for a respective plurality of theoretical signals in a store, each theoretical signal having its own particular attributes, the theoretical covariance matrices being for use in computing beamformer coefficients for application by a beamformer to signals received at the device;

receiving the signals at an input of the device;

determining, for the plurality of theoretical covariance matrices, a respective plurality of weights based on an analysis of the extent to which the received signals have the particular attributes of the theoretical signals;

retrieving the plurality of theoretical covariance matrices from the store;

calculating a covariance matrix set for the received signals by performing a weighted sum of the retrieved plurality of theoretical covariance matrices using the determined respective plurality of weights; and

computing beamformer coefficients based on the calculated covariance matrix for the received signals, said beamformer coefficients being for application by the beamformer to the received signals for generating a beamformer output.

2. The method of claim 1, further comprising determining the plurality of theoretical covariance matrices.

3. The method of claim 1, further comprising the beamformer applying the computed beamformer coefficients to the received signals, thereby generating the beamformer output.

4. The method of claim 1, wherein said particular attributes of a signal comprise a direction from which the signal is received at the input.

5. The method of claim 1, wherein said particular attributes of a signal comprise a distance from which the signal is received at the input.

6. The method of claim 1, wherein said particular attributes of a signal comprise a spectral shape of the signal.

7. The method of claim 6, wherein the spectral shape is that of white noise.

8. The method of claim 1, further comprising determining a mode in which the device is operating, wherein at least one of said determination of the plurality of weights is performed in dependence upon the determined mode, or the ones of the theoretical covariance matrices to be included in said weighted sum are selected based on the determined mode.

9. The method of claim 8, wherein the mode is determined based on characteristics of the received signals.

10. The method of claim 8, wherein the mode is determined based on an input from a user of the device.

11. The method of claim 8, wherein the mode is determined based on an analysis of signals that are output from the device.

12. The method of claim 1, wherein the calculating the covariance matrix for the received signals includes adding a normalizing data set to the weighted sum of the retrieved plurality of theoretical covariance matrices.

13. The method of claim 12, wherein the normalizing data set is a weighted identity matrix.

14. The method of claim 1, wherein the calculating the covariance matrix for the received signals is performed in dependence upon a determination that the beamformer coefficients are to be updated.

15. The method of claim 14, further comprising performing said determination that the beamformer coefficients are to be updated, said determination being based on characteristics of at least one of the received signals, or the beamformer output.

16. The method of claim 1, wherein said storing the plurality of theoretical covariance matrices in the store comprises storing the plurality of theoretical covariance matrices in a tabularized or parameterized form.

17. The method of claim 1, wherein said beamformer coefficients are computed so as to minimize the power in the beamformer output within beamforming constraints of the beamformer.

18. The method of claim 1, further comprising using the beamformer output to represent the signals received at the input for further processing within the device.

19. The method of claim 1, wherein the signals are one of audio signals, general broadband signals, general narrowband signals, radar signals sonar signals, antenna signals, radio waves, or microwaves.

20. A device for computing beamformer coefficients, the device comprising:

a store for storing a plurality of sets of theoretical covariance matrices for a respective plurality of theoretical signals, each theoretical signal having its own particular attributes, the theoretical covariance matrices being for use in computing beamformer coefficients for application by a beamformer to signals received at the device;

an input for receiving signals;

a processor; and

a memory comprising instructions executable by the processor to perform operations comprising;

determining, for the plurality of theoretical covariance matrices, a respective plurality of weights based on an analysis of the extent to which the received signals have the particular attributes of the theoretical signals;

retrieving the plurality of theoretical covariance matrices from the store;

calculating a covariance matrix for the received signals by performing a weighted sum of the retrieved plurality of theoretical covariance matrices using the determined respective plurality of weights; and

computing beamformer coefficients based on the calculated covariance matrix for the received signals, said beamformer coefficients being for application by the beamformer to the received signals for generating a beamformer output.

21. The device of claim 20, the operations further comprising determining the plurality of theoretical covariance matrices.

22. The device of claim 20, wherein the signals are audio signals and the input comprises a plurality of microphones for receiving the audio signals.

23. The device of claim 20, further comprising a beamformer configured to apply the computed beamformer coefficients to the received signals, to thereby generate the beamformer output.

24. The device of claim 23, wherein the beamformer is a Minimum Variance Distortionless Response beamformer.

27. The device of claim 20, wherein said particular attributes of a signal comprise a spectral shape of the signal.

28. The device of claim 27, wherein the spectral shape is that of white noise.

25. A computer program product for computing beamformer coefficients for signals received at an input of a device, the computer program product being embodied on a computer-readable memory device and configured so as when executed on a processor of the device to perform operations comprising:

storing a plurality of theoretical covariance matrices for a respective plurality of theoretical signals in a store, each theoretical signal having its own particular attributes, the theoretical covariance matrices being for use in computing beamformer coefficients for application by a beamformer to the signals received at the device;

determining, for the plurality of theoretical covariance matrices, a respective plurality of weights based on an analysis of the extent to which the received signals have the particular attributes of the theoretical signals;

retrieving the plurality of theoretical covariance matrices from the store;

calculating a covariance matrix for the received signals by performing a weighted sum of the retrieved plurality of theoretical covariance matrices using the determined respective plurality of weights; and

computing the beamformer coefficients based on the calculated covariance matrix for the received signals, said beamformer coefficients being for application by a beamformer to the received signals for generating a beamformer output.

26. A device configured to compute beamformer coefficients, the device comprising:

a store configured to store a plurality of theoretical covariance matrices for a respective plurality of theoretical signals, each theoretical signal having its own particular attributes, the covariance matrices being for use in computing beamformer coefficients for application by a beamformer to signals received at the device;

a receiver configured to receive signals;

a weight determining block configured to determine, for the plurality of theoretical covariance matrices, a respective plurality of weights based on an analysis of the extent to which the received signals have the particular attributes of the theoretical signals;

a retrieving block configured to retrieve the plurality of theoretical covariance matrices from the store;

a calculating block configured to calculate a covariance matrix for the received signals by performing a weighted sum of the retrieved plurality of theoretical covariance matrices using the determined respective plurality of weights; and

a computing block configured to compute the beamformer coefficients based on the calculated covariance matrix for the received signals, said beamformer coefficients being for application by the beamformer to the received signals for generating a beamformer output.