Processing audio signals using a discrete state machine

• US 5,592,547 A
• Filed: 12/20/1994
• Issued: 01/07/1997
• Est. Priority Date: 11/24/1993
• Status: Expired due to Term

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Ng, Seo P., Wu, Carol
• Primary Examiner(s): Brown, Thomas W.
• Assistant Examiner(s): SAINT SURIN, JACQUES M
• Application Number:
  US08/359,477
• Time in Patent Office: 749 Days
• Field of Search: 379/388, 379/420, 379/389, 379/390, 379/391, 395/325, 364/231.8
• US Class Current: 379/406.07
• CPC Class Codes: G06T 1/00 General purpose image data processing
• CPC Class Codes: G06T 2200/16 involving adaptation to the client's capabilities
• CPC Class Codes: H04L 12/1813 for computer conferences, e.g. chat rooms instant messaging H04L51/04; protocols for multimedia communication H04L65/1101; arrangements for multi-party communication H04L65/403; telephonic conference arrangements H04M3/56; television conference systems H04N7/15
• CPC Class Codes: H04M 3/567 Multimedia conference systems
• CPC Class Codes: H04M 3/568 audio processing specific to telephonic conferencing, e.g. spatial distribution, mixing of participants echo suppression in two-way loud-speaking telephone systems H04M9/02; sound field processing per se H04S7/30
• CPC Class Codes: H04M 7/006 Networks other than PSTN/ISDN providing telephone service, e.g. Voice over Internet Protocol (VoIP) , including next generation networks with a packet-switched transport layer H04L65/00 takes precedence; aspects not specific to the type of network H04M3/00; special services in those networks H04M3/42
• CPC Class Codes: H04M 9/08 Two-way loud-speaking telephone systems with means for conditioning the signal, e.g. for suppressing echoes for one or both directions of traffic
• CPC Class Codes: H04N 21/4143 embedded in a Personal Computer [PC]
• CPC Class Codes: H04N 21/643 Communication protocols network streaming protocols, e.g. real-time transport protocol [RTP] or real-time control protocol [RTCP], H04L65/65
• CPC Class Codes: H04N 7/148 Interfacing a video terminal to a particular transmission medium, e.g. ISDN
• CPC Class Codes: H04N 7/15 Conference systems
• CPC Class Codes: H04Q 2213/13003 Constructional details of switching devices
• CPC Class Codes: H04Q 2213/13103 Memory
• CPC Class Codes: H04Q 2213/13106 Microprocessor, CPU
• CPC Class Codes: H04Q 2213/13204 Protocols
• CPC Class Codes: H04Q 2213/13209 ISDN
• CPC Class Codes: H04Q 2213/1324 Conference call
• CPC Class Codes: H04Q 2213/13299 Bus
• CPC Class Codes: H04Q 2213/13337 Picturephone, videotelephony
• CPC Class Codes: H04Q 2213/13389 LAN, internet

Abstract

A level of transmit attenuation and a level of receive attenuation are generated using a discrete state machine of a controller. Local audio signals generated with a local microphone are attenuated with a transmit attenuator at the level of transmit attenuation for transmission to a remote node. Remote audio signals received from the remote node are attenuated with a receive attenuator at the level of receive attenuation for playback at a local speaker.

Claims

1. An apparatus for processing audio signals, comprising:

(a) a transmit attenuator for attenuating local audio signals generated by a local microphone for transmission to a remote node;

(b) a receive attenuator for attenuating remote audio signals received from the remote node for playback at a local speaker; and

(c) a controller, electrically connected to the transmit attenuator and the receive attenuator, for controlling a level of transmit attenuation for the transmit attenuator and a level of receive attenuation for the receive attenuator, wherein the controller comprises a discrete state machine to generate the level of transmit attenuation and the level of receive attenuation, wherein the discrete state machine comprises a no talking state, a single talk (local) state, a single talk (remote) state, and at least one double talk state.

2. The apparatus of claim 1, wherein the controller controls the sum of the level of transmit attenuation and the level of receive attenuation to correspond to unity gain.

3. The apparatus of claim 1, wherein the controller generates the levels of transmit and receive attenuation based on:

an energy level for the local audio signals;

an energy level for local background noise;

an energy level for the remote audio signals; and

an energy level for remote background noise.

4. The apparatus of claim 1, wherein:

the discrete state machine generates a continuous state variable i; and

the discrete state machine uses the continuous state variable i to generate the level of transmit attenuation and the level of receive attenuation.

5. The apparatus of claim 4, wherein the discrete state machine changes the continuous state variable i at two or more different rates, where the different rates correspond to different configurations of the local and remote audio signals.

6. The apparatus of claim 1, wherein the discrete state machine comprises the no talking state, the single talk (local) state, the single talk (remote) state, a double talk (local wins) state, and a double talk (remote wins) state.

7. The apparatus of claim 6, wherein:

if (T≦

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the no talking state;

if (T>

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the single talk (local) state;

if (T≦

τ

_L) &

(R>

τ

_R), then the discrete state machine is in the single talk (remote) state;

if (T>

τ

_L) &

(R>

τ

_R) &

[(T-τ

_L)-(R-τ

_R)]>

D1), then the discrete state machine is in the double talk (local wins) state; and

if (T>

τ

_L) &

(R>

τ

_R) &

[(R-τ

_R)-(L-τ

_L)]>

D2), then the discrete state machine is in the double talk (remote wins) state, wherein;

T is an energy for the local audio signals;

R is an energy for the remote audio signals;

τ

_L is a local background noise threshold;

τ

_R is a remote background noise threshold; and

D1 and D2 are specified threshold parameters.

8. The apparatus of claim 7, wherein state transitions for the discrete state machine comprise:

from the no talking state to either the single talk (local) state or the single talk (remote) state;

from the single talk (local) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the single talk (remote) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the double talk (local wins) state to only the single talk (local) state; and

from the double talk (remote wins) state to only the single talk (remote) state.

9. The apparatus of claim 8, wherein:

when the state transition is from the no talking state to the single talk (local) state, the discrete state machine increments a continuous state variable i by a fast step;

when the state transition is from the no talking state to the single talk (remote) state, the discrete state machine decrements the continuous state variable i by the fast step;

when the state transition is from the single talk (local) state to the no talking state, the discrete state machine decrements the continuous state variable i by a slow step;

when the state transition is from the single talk (local) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

when the state transition is from the single talk (local) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

when the state transition is from the single talk (remote) state to the no talking state, the discrete state machine increments the continuous state variable i by the slow step;

when the state transition is from the single talk (remote) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

when the state transition is from the single talk (remote) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

when the state transition is from the double talk (local wins) state to the single talk (local) state, the discrete state machine decrements the continuous state variable i by the slow step; and

when the state transition is from the double talk (remote wins) state to the single talk (remote) state, the discrete state machine increments the continuous state variable i by the slow step.

10. The apparatus of claim 9, wherein:

when the continuous state variable i is between -j and -k, the controller sets the level of transmit attenuation to 0 and the level of receive attenuation to 1, wherein;

-j≦

i≦

j;

j is a specified positive constant; and

k is a specified positive constant less than j;

when the continuous state variable i is between k and j, the controller sets the level of transmit attenuation to 1 and the level of receive attenuation to 0; and

when the continuous state variable i is between -k and k, the controller sets the level of transmit attenuation to α and

the level of receive attenuation to (1-α

), wherein 0≦

α

≦

1.

11. The apparatus of claim 10, wherein, when the continuous state variable i is between -k and k,α

=(i+k)/2k.

12. The apparatus of claim 1, wherein:

the discrete state machine comprises the no talking state, the single talk (local) state, the single talk (remote) state, a double talk (local wins) state, and a double talk (remote wins) state;

if (T≦

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the no talking state;

if (T>

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the single talk (local) state;

if (T≦

τ

_L) &

(R>

τ

_R), then the discrete state machine is in the single talk (remote) state;

if (T>

τ

_L) &

(R>

τ

_R) &

[(T-τ

_L)-(R-τ

_R)]>

D1), then the discrete state machine is in the double talk (local wins) state;

if (T>

τ

_L) &

(R>

τ

_R) &

[(R-τ

_R)-(L-τ

_L)]>

D2), then the discrete state machine is in the double talk (remote wins) state, wherein;

T is an energy for the local audio signals;

R is an energy for the remote audio signals;

τ

_L is a local background noise threshold;

τ

_R is a remote background noise threshold; and

D1 and D2 are specified threshold parameters;

state transitions for the discrete state machine comprise;

from the no talking state to either the single talk (local) state or the single talk (remote) state;

from the single talk (local) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the single talk (remote) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the double talk (local wins) state to only the single talk (local) state; and

from the double talk (remote wins) state to only the single talk (remote) state;

when the state transition is from the no talking state to the single talk (local) state, the discrete state machine increments a continuous state variable i by a fast step;

when the state transition is from the no talking state to the single talk (remote) state, the discrete state machine decrements the continuous state variable i by the fast step;

when the state transition is from the single talk (local) state to the no talking state, the discrete state machine decrements the continuous state variable i by a slow step;

when the state transition is from the single talk (local) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

when the state transition is from the single talk (local) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

when the state transition is from the single talk (remote) state to the no talking state, the discrete state machine increments the continuous state variable i by the slow step;

when the state transition is from the single talk (remote) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

when the state transition is from the single talk (remote) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

when the state transition is from the double talk (local wins) state to the single talk (local) state, the discrete state machine decrements the continuous state variable i by the slow step;

when the state transition is from the double talk (remote wins) state to the single talk (remote) state, the discrete state machine increments the continuous state variable i by the slow step;

when the continuous state variable i is between -j and -k, the controller sets the level of transmit attenuation to 0 and the level of receive attenuation to 1, wherein;

-j≦

i≦

j;

j is a specified positive constant; and

k is a specified positive constant less than j;

when the continuous state variable i is between k and j, the controller sets the level of transmit attenuation to 1 and the level of receive attenuation to 0; and

when the continuous state variable i is between -k and k, the controller sets the level of transmit attenuation to α and

the level of receive attenuation to (1-α

), wherein α

=(i+k)/2k.

24. The process of claim 1, wherein:

the discrete state machine comprises the no talking state, the single talk (local) state, the single talk (remote) state, a double talk (local wins) state, and a double talk (remote wins) state;

if (T≦

α

_L) &

(R≦

τ

_R), then the discrete state machine is in the no talking state;

if (T>

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the single talk (local) state;

if (T≦

τ

_L) &

(R>

τ

_R), then the discrete state machine is in the single talk (remote) state;

if (T>

τ

_L) &

(R>

τ

_R) &

[(T-τ

_L)-(R-τ

_R)]>

D1), then the discrete state machine is in the double talk (local wins) state;

if (T>

τ

_L) &

(R>

τ

_R) &

[(R-τ

_R)-(L-τ

_L)]>

D2), then the discrete state machine is in the double talk (remote wins) state, wherein;

T is an energy for the local audio signals;

R is an energy for the remote audio signals;

τ

_L is a local background noise threshold;

τ

_R is a remote background noise threshold; and

D1 and D2 are specified threshold parameters;

step (a) comprises the steps of;

(1) implementing state transitions by the discrete state machine, wherein the state transitions comprise;

from the no talking state to either the single talk (local) state or the single talk (remote) state;

from the single talk (local) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the single talk (remote) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the double talk (local wins) state to only the single talk (local) state; and

from the double talk (remote wins) state to only the single talk (remote) state;

(2) when the state transition is from the no talking state to the single talk (local) state, the discrete state machine increments a continuous state variable i by a fast step;

(3) when the state transition is from the no talking state to the single talk (remote) state, the discrete state machine decrements the continuous state variable i by the fast step;

(4) when the state transition is from the single talk (local) state to the no talking state, the discrete state machine decrements the continuous state variable i by a slow step;

(5) when the state transition is from the single talk (local) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

(6) when the state transition is from the single talk (local) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

(7) when the state transition is from the single talk (remote) state to the no talking state, the discrete state machine increments the continuous state variable i by the slow step;

(8) when the state transition is from the single talk (remote) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

(9) when the state transition is from the single talk (remote) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

(10) when the state transition is from the double talk (local wins) state to the single talk (local) state, the discrete state machine decrements the continuous state variable i by the slow step;

(11) when the state transition is from the double talk (remote wins) state to the single talk (remote) state, the discrete state machine increments the continuous state variable i by the slow step;

(12) when the continuous state variable i is between -j and -k, the controller sets the level of transmit attenuation to 0 and the level of receive attenuation to 1, wherein;

-j≦

i≦

j;

j is a specified positive constant; and

k is a specified positive constant less than j;

(13) when the continuous state variable i is between k and j, the controller sets the level of transmit attenuation to 1 and the level of receive attenuation to 0; and

(14) when the continuous state variable i is between -k and k, the controller sets the level of transmit attenuation to a and the level of receive attenuation to (1-α

), wherein α

=(i+k)/2k.

13. A computer-implemented process for processing audio signals, comprising the steps of:

(a) generating a level of transmit attenuation and a level of receive attenuation using a discrete state machine of a controller, wherein the discrete state machine comprises a no talking state, a single talk (local) state, a single talk (remote) state, and at least one double talk state;

(b) attenuating, with a transmit attenuator at the level of transmit attenuation, local audio signals generated with a local microphone for transmission to a remote node; and

(c) attenuating, with a receive attenuator at the level of receive attenuation, remote audio signals received from the remote node for playback at a local speaker.

14. The process of claim 13, wherein step (a) comprises the step of controlling the sum of the level of transmit attenuation and the level of receive attenuation to correspond to unity gain.

15. The process of claim 13, wherein step (a) comprises the step of generating the levels of transmit and receive attenuation based on:

an energy level for the local audio signals;

an energy level for local background noise;

an energy level for the remote audio signals; and

an energy level for remote background noise.

16. The process of claim 13, wherein step (a) comprises the steps of:

(1) generating a continuous state variable i with the discrete state machine; and

(2) using the continuous state variable i by the discrete state machine to generate the level of transmit attenuation and the level of receive attenuation.

17. The process of claim 16, wherein step (a)(1) comprises the step of changing the continuous state variable i at two or more different rates, where the different rates correspond to different configurations of the local and remote audio signals.

18. The process of claim 13, wherein the discrete state machine comprises the no talking state, the single talk (local) state, the single talk (remote) state, a double talk (local wins) state, and a double talk (remote wins) state.

19. The process of claim 18, wherein:

if (T≦

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the no talking state;

if (T>

τ

_L) &

(R≦

τ

_R), then the discrete state machine is in the single talk (local) state;

if (T≦

τ

_L) &

(R>

τ

_R), then the discrete state machine is in the single talk (remote) state;

if (T>

τ

_L) &

(R>

τ

_R) &

[(T-τ

_L)-(R-τ

_R)]>

D1), then the discrete state machine is in the double talk (local wins) state; and

if (T>

τ

_L) &

(R>

τ

_R) &

[(R-τ

_R)-(L-τ

_L)]>

D2), then the discrete state machine is in the double talk (remote wins) state, wherein;

T is an energy for the local audio signals;

R is an energy for the remote audio signals;

τ

_L is a local background noise threshold;

τ

_R is a remote background noise threshold; and

D1 and D2 are specified threshold parameters.

20. The process of claim 19, wherein step (a) comprises the step of implementing state transitions by the discrete state machine, wherein the state transitions comprise:

from the no talking state to either the single talk (local) state or the single talk (remote) state;

from the single talk (local) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the single talk (remote) state to either the no talking state, the double talk (local wins) state, or the double talk (remote wins) state;

from the double talk (local wins) state to only the single talk (local) state; and

from the double talk (remote wins) state to only the single talk (remote) state.

21. The process of claim 20, wherein step (a) comprises the steps of:

(1) when the state transition is from the no talking state to the single talk (local) state, the discrete state machine increments a continuous state variable i by a fast step;

(2) when the state transition is from the no talking state to the single talk (remote) state, the discrete state machine decrements the continuous state variable i by the fast step;

(3) when the state transition is from the single talk (local) state to the no talking state, the discrete state machine decrements the continuous state variable i by a slow step;

(4) when the state transition is from the single talk (local) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

(5) when the state transition is from the single talk (local) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

(6) when the state transition is from the single talk (remote) state to the no talking state, the discrete state machine increments the continuous state variable i by the slow step;

(7) when the state transition is from the single talk (remote) state to the double talk (local wins) state, the discrete state machine increments the continuous state variable i by the fast step;

(8) when the state transition is from the single talk (remote) state to the double talk (remote wins) state, the discrete state machine decrements the continuous state variable i by the fast step;

(9) when the state transition is from the double talk (local wins) state to the single talk (local) state, the discrete state machine decrements the continuous state variable i by the slow step; and

(10) when the state transition is from the double talk (remote wins) state to the single talk (remote) state, the discrete state machine increments the continuous state variable i by the slow step.

22. The process of claim 21, wherein:

step (a) further comprises the steps of;

(11) when the continuous state variable i is between -j and -k, the controller sets the level of transmit attenuation to 0 and the level of receive attenuation to 1, wherein;

-j≦

i≦

j;

j is a specified positive constant; and

k is a specified positive constant less than j;

(12) when the continuous state variable i is between k and j, the controller sets the level of transmit attenuation to 1 and the level of receive attenuation to 0; and

(13) when the continuous state variable i is between -k and k, the controller sets the level of transmit attenuation to α and

the level of receive attenuation to (1-α

), wherein 0≦

α

≦

1.

23. The process of claim 22, wherein, when the continuous state variable i is between -k and k,α

=(i+k)2k.