Fast echo canceller reconvergence after TDM slips and echo level changes

1. A method of adjusting an echo canceller, said method comprising:

obtaining a first cross-correlation between a far-end signal and an error signal, wherein said error signal is generated by subtracting an output signal of an adaptive filter from a local-end signal including an echo signal, wherein said adaptive filter utilizes said far-end signal to generate the output signal, and wherein said first cross-correlation is indicative of adequacy of cancellation of said echo signal;

determining whether said first cross-correlation is above a pro-determined threshold;

relocating said adaptive filter by a few samples if said determining determines that said first cross-correlation is above said pre-determined threshold indicative of said echo signal not being cancelled adequately;

calculating a first improvement indicator parameter, wherein said first improvement indicator parameter is calculated after said relocating said adaptive filter by said few samples; and

determining whether said first improvement indicator parameter indicates a performance improvement by said adaptive filter after said relocating said adaptive filter by said few samples;

wherein said first cross-correlation is defined by:

C

=



∑

0

N

-

1

⁢

⁢

e

⁡

(

n

)

·

x

⁡

(

n

-

k

)



∑

0

N

-

1

⁢

⁢

e

2

⁡

(

n

)

⁢

∑

0

N

-

1

⁢

⁢

x

2

⁡

(

n

-

k

)

⁢

⁢

where

,

0

<

C

<

1

wherein C designates said first cross-correlation, x(n) designates said far-end signal, e(n) designates said error signal.

2. The method of claim 1 , wherein said obtaining further includes calculating a first echo return loss enhancement (ERLE), and wherein said improvement indicator parameter is a second ERLE, and wherein said determining includes comparing said first ERLE with said second ERLE.

3. The method of claim 1 , wherein said first improvement indicator parameter is a second cross-correlation between said far-end signal and said error signal, and wherein said determining includes comparing said first cross-correlation with said second cross-correlation.

4. The method of claim 1 further comprising:

enabling aggressive adaption on of said adaptive filter if said determining does not determine said performance improvement.

5. The method of claim 1 , wherein said relocating relocates said adaptive filter forward by said few samples.

6. The method of claim 3 further comprising:

relocating said adaptive filter by a few samples backward if said determining does not determine said performance improvement.

7. The method of claim 6 further comprising:

calculating a second improvement indicator parameter, wherein said second improvement indicator parameter is calculated after said relocating said adaptive filter backward by said few samples; and

determining whether said second improvement indicator parameter indicates said performance improvement by said adaptive filter after said relocating said adaptive filter backward by said few samples.

8. The method of claim 7 further comprising:

enabling aggressive adaptation of said adaptive filter if said determining does not determine said performance improvement by said adaptive filter after said relocating said adaptive filter backward by said few samples.

9. The method of claim 7 further comprising:

calculating a gain based on said local-end signal and said error signal if said determining does not determine said performance improvement by said adaptive filter after said relocating said adaptive filter backward by said few samples; and

multiplying said adaptive filter by said gain.

10. The method of claim 9 further comprising:

calculating a third improvement indicator parameter, wherein said third improvement indicator parameter is calculated after said multiplying said adaptive filter by said gain;

determining whether said third improvement indicator parameter indicates said performance improvement by said adaptive filter after said multiplying said adaptive filter by said gain; and enabling aggressive adaptation of said adaptive filter if said determining does not determine said performance improvement by said adaptive filter after said multiplying said adaptive filter by said gain.

11. The method of claim 1 , wherein said relocating relocates said adaptive filter backward by said few samples.

12. The method of claim 1 further comprising:

calculating a gain based on said local-end signal and said error signal if said determining does not determine said performance improvement; and

multiplying said adaptive filter by said gain.

13. The method of claim 12 further comprising:

calculating a second improvement indicator parameter, wherein said second improvement indicator parameter is calculated after said multiplying said adaptive filter by said gain;

determining whether said second improvement indicator parameter indicates said performance improvement by said adaptive filter after said multiplying said adaptive filter by said gain; and

enabling aggressive adaptation of said adaptive filter if said determining does not determine said performance improvement by said adaptive filter after said multiplying said adaptive filter by said gain.

14. A method of adjusting an echo canceller, said method comprising:

obtaining a first cross-correlation between a far-end signal and an error signal, wherein said error signal is generated by subtracting an output signal of an adaptive filter from a local-end signal including an echo signal, wherein said adaptive filter utilizes said far-end signal to generate the output signal, and wherein said first cross-correlation is indicative of adequacy of cancellation of said echo signal;

determining whether said first cross-correlation is above a pre-determined threshold;

calculating a gain based on said local-end signal and said error signal if said determining determines that said first cross-correlation is above said pre-determined threshold indicative of said echo signal not being cancelled adequately;

multiplying said adaptive filter by said gain;

calculating a first improvement indicator parameter, wherein said first improvement indicator parameter is calculated after said multiplying said adaptive filter by said gain; and

determining whether said first improvement indicator parameter indicates a performance improvement by said adaptive filter after said multiplying said adaptive filter by said gain;

wherein said first cross-correlation is defined by:

C

=



∑

0

N

-

1

⁢

⁢

e

⁡

(

n

)

·

x

⁡

(

n

-

k

)



∑

0

N

-

1

⁢

⁢

e

2

⁡

(

n

)

⁢

∑

0

N

-

1

⁢

⁢

x

2

⁡

(

n

-

k

)

⁢

⁢

where

,

0

<

C

<

1

wherein C designates said first cross-correlation, x(n) designates said far-end signal, e(n) designates said error signal.

15. The method of claim 14 , wherein said obtaining further includes calculating a first echo return loss enhancement (ERLE), and wherein said improvement indicator parameter is a second ERLE, and wherein said determining includes comparing said first ERLE with said second ERLE.

16. The method of claim 14 , wherein said first improvement indicator parameter is a second cross-correlation between said far-end signal and said error signal, and wherein said determining includes comparing said first cross-correlation with said second cross-correlation.

17. The method of claim 14 further comprising:

enabling aggressive adaptation of said adaptive filter if said determining does not determine said performance improvement.

18. The method of claim 14 further comprising:

relocating said adaptive filter by a few samples if said determining does not determine said performance improvement.

19. The method of claim 18 , wherein said relocating relocates said adaptive filter forward by said few samples.

20. The method of claim 18 , wherein said relocating relocates said adaptive filter backward by said few samples.

21. An echo canceller comprising:

a cross-correlator configured to obtain a first cross-correlation between a far-end signal and an error signal, wherein said error signal is generated by subtracting an output signal of an adaptive filter from a local-end signal including an echo signal, wherein said adaptive filter utilizes said far-end signal to generate the output signal, and wherein said first cross-correlation is indicative of adequacy of cancellation of said echo signal;

an adaptive filter configured to determine whether said first cross-correlation is above a pre-determined threshold, said adaptive filter having a relocation module configured to relocate said adaptive filter by a few samples if said adaptive filter determines that said first cross-correlation is above said pre-determined threshold indicative of said echo signal not being cancelled adequately;

wherein said echo canceller is configured to calculate a first improvement indicator parameter, wherein said first improvement indicator parameter is calculated after said relocating said adaptive filter by said few samples, and wherein said echo canceller is configured to determine whether said first improvement indicator parameter indicates a performance improvement by said adaptive filter after said relocating said adaptive filter by said few samples, and wherein said first cross-correlation is defined by:

C

=



∑

0

N

-

1

⁢

⁢

e

⁡

(

n

)

·

x

⁡

(

n

-

k

)



∑

0

N

-

1

⁢

⁢

e

2

⁡

(

n

)

⁢

∑

0

N

-

1

⁢

⁢

x

2

⁡

(

n

-

k

)

⁢

⁢

where

,

0

<

C

<

1

wherein C designates said first cross-correlation, x(n) designates said far-end signal, e(n) designates said error signal.

22. The echo canceller of claim 21 , wherein said echo canceller obtains a first echo return loss enhancement (ERLE), and wherein said improvement indicator parameter is a second ERLE, and wherein said echo canceller compares said first ERLE with said second ERLE to determine said performance improvement.

23. The echo canceller of claim 21 , wherein said first improvement indicator parameter is a second cross-correlation between said far-end signal and said error signal, and wherein said echo canceller compares said first cross-correlation with said second cross-correlation to determine said performance improvement.

24. The echo canceller of claim 21 , wherein said relocation module relocates said adaptive filter forward by said few samples.

25. The echo canceller of claim 21 , wherein said relocation module relocates said adaptive filter backward by said few samples.

26. The echo canceller of claim 21 , wherein an aggressive adaptation of said adaptive filter is enabled if said echo canceller does not determine said performance improvement by said adaptive filter.

27. The echo canceller of claim 21 further comprising a level adjustment module configured to calculate a gain based on said local-end signal and said error signal if said echo canceller does not determine said performance improvement by said adaptive filter, and further configured to multiply said adaptive filter by said gain.

28. The echo canceller of claim 27 , wherein said echo canceller calculates a second improvement indicator parameter, wherein said second improvement indicator parameter is calculated after said level adjustment module multiplies said adaptive filter by said gain, wherein said echo canceller is configured to determine whether said second improvement indicator parameter indicates said performance improvement by said adaptive filter after said level adjustment module multiplies said adaptive filter by said gain, and wherein an aggressive adaptation of said adaptive filter is enabled if said echo canceller does not determine said performance improvement by said adaptive filter after said level adjustment module multiplies said adaptive filter by said gain.