Apparatus and method for canceling acoustic echoes including non-linear distortions in loudspeaker telephones

• US 5,680,450 A
• Filed: 02/24/1995
• Issued: 10/21/1997
• Est. Priority Date: 02/24/1995
• Status: Expired due to Fees

• Current Assignee: Ericsson, Inc. (Telefonaktiebolaget LM Ericsson)
• Original Assignee: Ericsson, Inc. (Telefonaktiebolaget LM Ericsson)
• Inventors: Dent, Paul W., Solve, Torbjorn W.
• Primary Examiner(s): Zele, Krista M.
• Assistant Examiner(s): Kumar, Devendra
• Application Number:
  US08/393,711
• Time in Patent Office: 970 Days
• Field of Search: 379/388, 379/389, 379/390, 379/420, 379/406, 379/410, 379/411, 379/409, 381/66, 381/59, 381/96, 381/98, 381/103, 370/32.1, 370/286, 370/289, 370/290, 370/291
• US Class Current: 379/406.08
• CPC Class Codes: H04M 9/082 using echo cancellers echo cancellers per se H04B3/23

Abstract

An echo canceling loudspeaker telephone includes a loudspeaker which produces a sound pressure wave in response to an input signal which is applied to an audio input thereof. This sound pressure wave includes a desired linear component which is a linear function of the input signal, and an undesired non-linear component which is a non-linear function of the input signal, and the sound pressure wave is transmitted along an acoustic path. A microphone is positioned in the acoustic path and converts the sound pressure wave into an output signal. An echo filter is responsive to the input signal and generates an estimated echo signal. This echo filter includes a loudspeaker model which generates an estimate of the sound pressure wave including an estimate of the linear component and an estimate of the non-linear component. This echo filter also includes an acoustic path model which generates an estimate of the acoustic path from the loudspeaker to the microphone. In addition, a subtractor subtracts the estimated echo signal from the output signal thereby reducing an echo portion of said sound signal.

Claims

1. An echo canceling loudspeaker telephone comprising:

a loudspeaker for producing a sound pressure wave in response to an input signal which is applied to an audio input thereof, said sound pressure wave including a desired linear component which is a linear function of said input signal, and an undesired non-linear component which is a non-linear function of said input signal, said sound pressure wave being transmitted along an acoustic path;

a microphone positioned in said acoustic path for converting said sound pressure wave into an output signal;

an echo filter responsive to said input signal and which generates an estimated echo signal, said echo filter comprising a loudspeaker model which generates an estimate of said sound pressure wave including an estimate of said linear component and an estimate of said non-linear component, and an acoustic path model which generates an estimate of said acoustic path from said loudspeaker to said microphone;

a subtractor for subtracting said estimated echo signal from said output signal thereby reducing an echo portion of said output signal; and

an echo filter modifier responsive to said output signal said input signal and said estimated echo signal which determines a residual echo portion of said output signal remaining after subtracting said estimated echo signal from said output signal, and which modifies estimates of distortions due to said loudspeaker and said acoustic path in response to said residual echo portion to further reduce said echo portion of said output signal;

wherein said loudspeaker model and said acoustic path model are connected in series.

2. An echo-canceling loudspeaker telephone according to claim 1 wherein said echo filter modifier comprises a loudspeaker model modifier for modifying the estimate of said sound pressure wave including said estimate of said linear component and said estimate of said non-linear component.

3. An echo-canceling loudspeaker telephone according to claim 1 wherein said echo filter modifier comprises an acoustic path model modifier for improving the estimate of said acoustic path from said loudspeaker to said microphone.

4. An echo-canceling loudspeaker telephone according to claim 1 wherein said acoustic path model comprises a finite-impulse-response filter.

5. An echo-canceling loudspeaker telephone according to claim 1 wherein said echo filter comprises a digital signal processor.

6. An echo-canceling loudspeaker telephone according to claim 1 wherein said loudspeaker model comprises means for performing a transformation of said input signal, said transformation being an estimate of a transfer function of said loudspeaker, said transfer function including a non-linear component.

7. An echo-canceling loudspeaker telephone according to claim 6 wherein said non-linear component represents one of a delay modulation of said loudspeaker and a diaphragm stress-strain curve of said loudspeaker.

8. An echo canceling loudspeaker telephone comprising:

a loudspeaker for producing a sound pressure wave in response to an input signal which is applied to an audio input thereof, said sound pressure wave including a desired linear component which is a linear function of said input signal, and an undesired non-linear component which is a non-linear function of said input signal, said sound pressure wave being transmitted along an acoustic path;

a microphone positioned in said acoustic path for converting said sound pressure wave into an output signal;

an echo filter responsive to said input signal and which generates an estimated echo signal, said echo filter comprising a loudspeaker model which generates an estimate of said sound pressure wave including an estimate of said linear component and an estimate of said non-linear component, and an acoustic path model comprising a first processing block for generating an estimate of said acoustic path for said non-linear component of said sound pressure wave and a second processing block for generating an estimate of said acoustic path for said linear component of said sound pressure wave; and

a subtractor for subtracting said estimated echo signal from said output signal thereby reducing an echo portion of said sound signal.

9. An echo-canceling loudspeaker telephone according to claim 8 further comprising an echo filter modifier responsive to said output signal, said input signal and said estimated echo signal for modifying said echo filter to further reduce said echo portion of said output signal.

10. An echo-canceling loudspeaker telephone according to claim 9 wherein said echo filter modifier comprises an loudspeaker model modifier for modifying the estimate of said sound pressure wave including said estimate of said linear component and said estimate of said non-linear component.

11. An echo-canceling loudspeaker telephone according to claim 9 wherein said echo filter modifier comprises an acoustic path model modifier for improving the estimate of said acoustic path from said loudspeaker to said microphone.

12. An echo-canceling loudspeaker telephone according to claim 8 wherein said acoustic path model comprises a finite-impulse-response filter.

13. An echo-canceling loudspeaker telephone according to claim 8 wherein said echo filter comprises a digital signal processor.

14. An echo-canceling loudspeaker telephone according to claim 8 wherein said loudspeaker model comprises means for performing a transformation of said input signal, said transformation being an estimate of a transfer function of said loudspeaker, said transfer function including a non-linear component.

15. An echo-canceling loudspeaker telephone according to claim 14 wherein said non-linear component represents one of a delay modulation of said loudspeaker and a diaphragm stress-strain curve of said loudspeaker.

16. An echo-canceling loudspeaker telephone comprising:

output transducer means for producing a sound pressure wave in response to an input signal which is applied to an audio input thereof, said sound pressure wave including a desired linear component which is a linear function of said input signal, and an undesired non-linear component which is a non-linear function of said input signal, said sound pressure wave being transmitted along an acoustic path;

input transducer means positioned in said acoustic path for converting said sound pressure wave into a output signal;

echo filter means responsive to said input signal for generating an estimated echo signal, said echo filter means comprising an output transducer model for generating an estimate of said sound pressure wave including an estimate of said linear component and an estimate of said non-linear component, and an acoustic path model for generating an estimate of said acoustic path from said output transducer means to said input transducer means;

combination means for combining said estimated echo signal and said output signal thereby reducing an echo portion of said output signal; and

an echo filter modifier responsive to said output signal, said input signal and said estimated echo signal which determines a residual echo portion of said output signal remaining after subtracting said estimated echo signal from said output signal, and which modifies estimates of distortions due to said loudspeaker and said acoustic path in response to said residual echo portion to further reduce said echo portion of said output signal.

17. An echo-canceling loudspeaker telephone according to claim 16 wherein said echo filter modifier comprises an output transducer model modifier for modifying the estimate of said sound pressure wave including said estimate of said linear component and said estimate of said non-linear component.

18. An echo-canceling loudspeaker telephone according to claim 16 wherein said echo filter modifier comprises an acoustic path model modifier for improving the estimate of said acoustic path from said output transducer to said input transducer.

19. An echo-canceling loudspeaker telephone according to claim 16 wherein said acoustic path model comprises a finite-impulse-response filter.

20. An echo-canceling loudspeaker telephone according to claim 16 wherein said echo filter means comprises a digital signal processor.

21. An echo-canceling loudspeaker telephone according to claim 16 wherein said output transducer means comprises a loudspeaker.

22. An echo-canceling loudspeaker telephone according to claim 16 wherein said input transducer means comprises a microphone.

23. An echo-canceling loudspeaker telephone according to claim 16 wherein said output transducer model and said acoustic path model are connected in series between said audio input and said combination means.

24. An echo-canceling loudspeaker telephone according to claim 16 wherein said combination means comprises a subtractor for subtracting said estimated echo signal from said output signal.

25. An echo-canceling loudspeaker telephone according to claim 16 wherein said output transducer model comprises means for performing a transformation of said input signal, said transformation being an estimate of a transfer function of said output transducer means, said transfer function including a non-linear component.

26. An echo-canceling loudspeaker telephone according to claim 25 wherein said non-linear component represents one of a delay modulation of said output transducer means and a diaphragm stress-strain curve of said output transducer means.

27. An echo-canceling loudspeaker telephone comprising:

output transducer means for producing a sound pressure wave in response to an input signal which is applied to an audio input thereof, said sound pressure wave including a desired linear component which is a linear function of said input signal, and an undesired non-linear component which is a non-linear function of said input signal, said sound pressure wave being transmitted along an acoustic path;

input transducer means positioned in said acoustic path for converting said sound pressure wave into a output signal;

echo filter means responsive to said input signal for generating an estimated echo signal, said echo filter means comprising an output transducer model for generating an estimate of said sound pressure wave including an estimate of said linear component and an estimate of said non-linear component, and an acoustic path model for generating an estimate of said acoustic path from said output transducer means to said input transducer means; and

combination means for combining said estimated echo signal and said output signal thereby reducing an echo portion of said output signal;

wherein said acoustic path model comprises a first processing block for generating an estimate of said acoustic path for said non-linear component of said sound pressure wave and a second processing block for generating an estimate of said acoustic path for said linear component of said sound pressure wave.

28. A method for reducing echoes in a telephone system including a loudspeaker, said method comprising the steps of:

applying an input signal to said loudspeaker to produce a sound pressure wave which is transmitted along an acoustic path, said sound pressure wave including a desired linear component which is a linear function of said input signal and an undesired non-linear component which is a non-linear function of said input signal;

converting said sound pressure wave in said acoustic path to produce an output signal including an echo portion;

generating an estimated echo signal in response to said input signal, wherein said estimated echo signal includes an estimate of distortions due to said loudspeaker which includes linear and non-linear components, and an estimate of distortions due to said acoustic path; and

combining said output signal and said estimated echo signal to reduce said echo portion of said output signal;

wherein said combining step is followed by the steps of determining a residual echo portion of said output signal remaining after said combining step, and modifying said estimates of distortions due to said loudspeaker and said acoustic path in response to said residual echo portion to further reduce said echo portion of said output signal.

29. A method according to claim 28 further comprising the step of comparing said input signal with said output signal to determine when said output signal substantially comprises only said echo portion, and wherein said modifying step is performed when said output signal substantially comprises only said echo portion.

30. A method according to claim 29 wherein said input signal comprises an input test signal.

31. An echo canceling loudspeaker telephone comprising:

a loudspeaker for producing a sound pressure wave in response to an input signal which is applied to an audio input thereof, said sound pressure wave including a desired linear component which is a linear function of said input signal, and an undesired non-linear component which is a non-linear function of said input signal, said sound pressure wave being transmitted along an acoustic path;

a microphone positioned in said acoustic path for converting said sound pressure wave into an output signal;

an echo filter responsive to said input signal and which generates an estimated echo signal, said echo filter comprising a loudspeaker model which generates an estimate of said sound pressure wave including an estimate of said linear component and an estimate of said non-linear component, and an acoustic path model which generates an estimate of said acoustic path from said loudspeaker to said microphone; and

a subtractor for subtracting said estimated echo signal from said output signal thereby reducing an echo portion of said output signal;

wherein said loudspeaker model and said acoustic path model are connected in series; and

wherein said acoustic path model comprises a first processing block for generating an estimate of said acoustic path for said non-linear component of said sound pressure wave and a second processing block for generating an estimate of said acoustic path for said linear component of said sound pressure wave.