Systems and methods of echo reduction

1. A method of echo cancellation, the method comprising:

producing, by a loudspeaker, acoustic waves based on a far-microphone signal;

receiving, at a local microphone, an echo based on the acoustic waves, and receiving acoustic waves generated locally, the receiving creates a local-microphone signal;

producing an estimated-echo signal based on the far-microphone signal and a step-size parameter;

summing the local-microphone signal and the estimated-echo signal to produce a resultant signal having reduced echo in relation to the local-microphone signal; and

controlling the step-size parameter by:

calculating a convergence value based on a cross-correlation of the local-microphone signal and the resultant signal; and

updating the step-size parameter based on the convergence value,

wherein calculating the convergence value further comprises:

converting the local-microphone signal into the frequency domain to create a local-microphone spectrum;

converting the resultant signal into the frequency domain to create a resultant spectrum;

performing a cross-correlation of the local-microphone spectrum and the resultant spectrum to create a cross-correlation spectrum; and

creating the convergence value based on the cross-correlation spectrum.

2. The method of claim 1 wherein updating the step-size parameter further comprises decreasing the current step-size parameter relative to a previous step-size parameter when the convergence value is larger than a predetermined upper value.

3. The method of claim 1 wherein updating the step-size parameter further comprises increasing the current step-size parameter relative to a previous step-size parameter when the convergence value is smaller than a predetermined lower value.

4. The method of claim 1 wherein updating the convergence value further comprises:

calculating a variance of the local-microphone spectrum;

dividing the cross-correlation spectrum by the variance to create a dividend spectrum; and

subtracting the dividend spectrum from one to create a convergence spectrum; and

calculating the convergence value as an average amplitude of members of the convergence spectrum.

5. The method of claim 1 wherein updating the step-size parameter further comprises at least one selected from a group comprising: multiplying a previous step-size parameter by a predetermined value; and dividing the previous step-size parameter by the predetermined value.

6. The method of claim 1 wherein updating the step-size parameter further comprises at least one selected from a group comprising:

dividing a previous step-size parameter by a predetermined value to produce a quotient, and subtracting the quotient from the previous step-size parameter to produce the current step-size parameter; and

dividing a previous step-size parameter by a predetermined value to produce a quotient, and adding the quotient to the previous step-size parameter to produce the current step-size parameter.

7. An audio processing device comprising:

an audio-input port, an audio-output port, a microphone port, and a resultant port;

an echo cancellation controller defining an audio input coupled to the audio-input port, an audio output coupled the audio-output port, a microphone input coupled to the microphone port, a resultant output coupled to the resultant port, and a step-size input;

the echo cancellation controller is configured to reduce an echo received on the microphone input based on an audio signal received on the audio-input port and a step-size parameter received on the step-size input;

a step-size controller defining a first input coupled to the microphone port, a second input coupled to the resultant output, and a step-size output coupled to the step-size input of the echo cancellation controller;

the step-size controller configured to drive the step-size parameter to the step-size input by being configured to:

calculate a cross-correlation of a microphone signal received on the microphone input and a resultant signal received from the resultant output;

create a convergence value based on the cross-correlation; and

update the step-size parameter based on the convergence value,

wherein the step-size controller further comprises:

a Fourier-transform logic defining a first input coupled to the microphone input, a second input coupled to the resultant output, a microphone output, and a resultant output;

the Fourier-transform logic configured to generate a microphone spectrum from the microphone signal and to drive the microphone spectrum to the microphone output, and the Fourier-transform logic configured to generate a resultant spectrum from the resultant signal and drive the resultant spectrum to the resultant output of the Fourier-transform logic;

a cross-correlation logic having a first input coupled to the microphone output, and a second input coupled to the resultant output, and a convergence output;

the cross-correlation logic configured to perform a cross-correlation of the microphone spectrum and the resultant spectrum to create a cross-correlation spectrum; and

the cross-correlation logic configured to create the convergence value based on the cross-correlation spectrum, and drive the convergence value to the convergence output.

8. The audio processing device of claim 7 further comprising:

a step-size logic defining a convergence input coupled to convergence output, and the step-size output; and

the step-size logic configured to update the step-size parameter by decreasing the step-size parameter relative to a previous step-size parameter when the convergence value is larger than a predetermined upper value.

9. The audio processing device of claim 7 further comprising:

a step-size logic defining a convergence input coupled to convergence output, and the step-size output; and

the step-size logic configured to update the step-size parameter by increasing the step-size parameter relative to a previous step-size parameter when the convergence value is smaller than a predetermined lower value.

10. The audio processing device of claim 7 further comprising the cross-correlation logic defines a variance output, and the cross-correlation logic is configured to

calculate a variance of the cross-correlation spectrum;

divide the cross-correlation spectrum by the variance to create a dividend spectrum;

subtract the dividend spectrum from one to create a convergence spectrum; and

calculate the convergence value as an average amplitude of members of the convergence spectrum.

11. The audio processing device of claim 7 wherein when the step-size controller updates the step-size parameter, the step-size controller is further configured to at least one selected from a group comprising: multiply a previous step-size parameter by a predetermined value; and divide the previous step-size parameter by the predetermined value.

12. The audio processing device of claim 7 wherein when the step-size controller updates the step-size parameter, the step-size controller is further configured to at least one selected from a group comprising:

divide a previous step-size parameter by a predetermined value to produce a quotient, and subtracting the quotient from the previous step-size parameter to produce the step-size parameter; and

divide a previous step-size parameter by a predetermined value to produce a quotient, and adding the quotient to the previous step-size parameter to produce the step-size parameter.

13. An audio system comprising:

a loudspeaker;

a microphone;

an audio processing device comprising

an audio-input port, and a resultant port;

an echo cancellation controller defining an audio input coupled to the audio-input port, an audio output coupled the loudspeaker, a microphone input coupled to the microphone, a resultant output coupled to the resultant port, and a step-size input;

the echo cancellation controller is configured to reduce an echo received on the microphone based on an audio signal received on the audio-input port and a step-size parameter received on the step-size input;

a step-size controller defining a microphone input coupled to the microphone, a resultant input coupled to the resultant output, and a step-size output coupled to the step-size input of the echo cancellation controller;

the step-size controller configured to drive the step-size parameter to the step-size input by being configured to:

calculate a cross-correlation of a microphone signal received from the microphone and a resultant signal received on the resultant input;

create a convergence value based on the cross-correlation; and

update the step-size parameter based on the convergence value,

wherein the step-size controller further comprises:

a Fourier-transform logic defining a first input coupled to the microphone, a second input coupled to the resultant output, a microphone output, and a resultant output;

the Fourier-transform logic configured to generate a microphone spectrum from the microphone signal and to drive the microphone spectrum to the microphone output, and the Fourier-transform logic configured to generate a resultant spectrum from the resultant signal and drive the resultant spectrum to the resultant output of the Fourier-transform logic;

a cross-correlation logic having a first input coupled to the microphone output, and a second input coupled to the resultant output, and a convergence output;

the cross-correlation logic configured to perform a cross-correlation of the microphone spectrum and the resultant spectrum to create a cross-correlation spectrum; and

the cross-correlation logic configured to create the convergence value based on the cross-correlation spectrum, and drive the convergence value to the convergence output.

14. The audio system of claim 13 further comprising:

a step-size logic defining a convergence input coupled to the convergence output, and the step-size output; and

the step-size logic configured to update the step-size parameter by decreasing the step-size parameter relative to a previous step-size parameter when the convergence value is larger than a predetermined upper value.

15. The audio system of claim 13 further comprising:

a step-size logic defining a convergence input coupled to convergence output, and the step-size output; and

the step-size logic configured to update the step-size parameter by increasing the step-size parameter relative to a previous step-size parameter when the convergence value is smaller than a predetermined lower value.

16. The audio system of claim 13 further comprising the cross-correlation logic defines a variance output, and the cross-correlation logic is configured to

calculate a variance of the cross-correlation spectrum;

divide the cross-correlation spectrum by the variance to create a dividend spectrum; and

subtract the dividend spectrum from one to create a convergence spectrum; and

calculate the convergence value as an average amplitude of members of the convergence spectrum.

17. The audio system of claim 13 wherein when the step-size controller updates the step-size parameter, the step-size controller is further configured to at least one selected from a group comprising: multiply a previous step-size parameter by a predetermined value; and divide the previous step-size parameter by the predetermined value.