Half-duplex speakerphone

Claim 2

A speakerphone including a transmit audio path for conveying a transmit signal including speech signals and a receive audio path for conveying a receive signal including speech signals, the speakerphone comprising: control means for controlling an operation of said transmit audio path and said receive audio path, said control means having a plurality of input signals and a plurality of output signals, a value of said output signals being a function of a value of said input signals, said output signals including at least a signal expressive of, when asserted, a transmit state, a signal expressive of, when asserted, a receive state and a signal expressive of, when asserted, an idle state, said input signals including a first input signal having a value which is a function of: a transmit signal power estimate, a transmit audio path noise signal estimate and a transmit audio path echo signal estimate; and a receive signal power estimate, a receive audio path noise signal estimate and a receive audio path echo signal estimate.

Claim 3

A speakerphone as set forth in claim 1 wherein said input signals further include: a second input signal having a value which is a function of the estimate of the power of the transmit signal and the estimate of the noise signal for the transmit audio path, the second signal indicating if the transmit audio path is conveying speech signals; and a third input signal having a value which is a function of the estimate of the power of the receive signal and the estimate of the noise signal for the receive audio path, the third signal indicating if the receive audio path is conveying speech signals.

Claim 4

A speakerphone as set forth in claim 2 wherein said input signals further include: a fourth input signal which is expressive of whether said speakerphone is in a mute condition.

Claim 5

A speakerphone as set forth in claim 2 wherein said second input signal is generated by means comprising: means for multiplying said transmit signal power estimate by a first multiplication factor to obtain a scaled transmit signal power estimate signal; means for subtracting from said scaled transmit signal power estimate signal said transmit audio path noise signal; and means for subtracting from said transmit signal power estimate signal a first subtracting means signal having a value which is related to a desired transmit signal signal to noise ratio.

Claim 6

A speakerphone as set forth in claim 4 wherein said third input signal is generated by means comprising: means for multiplying said receive signal power estimate by a second multiplication factor to obtain a scaled receive signal power estimate signal; means for subtracting from said scaled receive signal power estimate signal said receive audio path noise signal; and means for subtracting from said receive signal power estimate signal a second subtracting means signal having a value which is related to a desired receive signal signal to noise ratio.

Claim 7

A speakerphone as set forth in claim 5 wherein said first and said second multiplication factor are each equal to approximately 0.75 and wherein said first and said second subtracting means signals each have a value of approximately 0.00305.

Claim 8

A speakerphone as set forth in claim 1 wherein said transmit signal power estimate signal and said receive signal power estimate signal are each generated by an envelope detection means having an associated rise time and an associated fall time.

Claim 9

A speakerphone as set forth in claim 7 wherein said rise time is approximately one millisecond and wherein said fall time is approximately 28 milliseconds.

Claim 10

A speakerphone as set forth in claim 1 wherein said transmit audio path echo estimate signal and said receive audio path echo estimate signal are each generated by an envelope detection means having an associated rise time and an associated fall time.

Claim 11

A speakerphone as set forth in claim 9 wherein said rise time is approximately 0.25 millisecond and wherein said fall time is approximately 100 milliseconds.

Claim 12

A speakerphone as set forth in claim 1 wherein said control means output signals further comprise a transmit audio path gain defining signal and a receive audio path gain defining signal and wherein said gain defining signals vary the gain of each of said transmit and said receive audio paths within a range of approximately 0 dB to less than approximately -50 dB when either said signal expressive of a transmit state is asserted or said signal expressive of a receive state is asserted.

Claim 13

A speakerphone as set forth in claim 11 wherein said gain defining signals define the gain of said transmit audio path to be approximately -10 dB and the gain of said receive audio path to be less than approximately -40 dB when said signal expressive of an idle state is asserted.

Claim 14

A speakerphone as set forth in claim 12 wherein said control means comprises means for generating said gain defining signals such that said gain defining signals each have a substantially Gaussian waveshape.

Claim 15

A speakerphone including a transmit audio path for conveying, when active, a transmit signal including a speech signal and a receive audio path for conveying, when active, a receive signal including a speech signal, the speakerphone comprising: first means for generating an estimate of the power of the transmit signal; second means for generating an estimate of a noise signal component associated with the transmit audio path; third means for generating an estimate of an echo signal component associated with the transmit audio path; fourth means for generating an estimate of the power of the receive signal; fifth means for generating an estimate of a noise signal component associated with the receive audio path; sixth means for generating an estimate of an echo signal component associated with the receive audio path; first means for comparing the estimate of the power of the transmit signal to the estimate of the noise signal for the transmit audio path, the first comparing means having an output for indicating if the transmit audio path is conveying speech signals; second means for comparing the estimate of the power of the receive signal to the estimate of the noise signal for the receive audio path, the second comparing means having an output for indicating if the receive audio path is conveying speech signals; third means for comparing the estimate of the transmit audio path signal power, noise signal and echo signal with the estimate of the receive audio path signal power, noise signal and echo signal, the third comparing means having an output for indicating whether the transmit audio path or the receive audio path is active; and means, coupled to and responsive to the output of each of the first, the second and the third comparing means, for selecting at least a transmit state, wherein the transmit audio path is active, a receive state, wherein the receive audio path is active, or an idle state, wherein neither the transmit or the receive audio paths are active.

Claim 16

A speakerphone including a transmit audio path for conveying, when active, a transmit signal including a speech signal and a receive audio path for conveying, when active, a receive signal including a speech signal, the speakerphone comprising: first means for generating an estimate of the power of the transmit signal; second means for generating an estimate of a noise signal component associated with the transmit audio path; third means for generating an estimate of an echo signal component associated with the transmit audio path; fourth means for generating an estimate of the power of the receive signal; fifth means for generating an estimate of a noise signal component associated with the receive audio path; sixth means for generating an estimate of an echo signal component associated with the receive audio path; first means for comparing the estimate of the power of the transmit signal to the estimate of the noise signal for the transmit audio path, the first comparing means having an output for indicating if the transmit audio path is conveying speech signals; second means for comparing the estimate of the power of the receive signal to the estimate of the noise signal for the receive audio path, the second comparing means having an output for indicating if the receive audio path is conveying speech signals; third means for comparing the estimate of the transmit audio path signal power, noise signal and echo signal with the estimate of the receive audio path signal power, noise signal and echo signal, the third comparing means having an output for indicating whether the transmit audio path or the receive audio path is active; and means, coupled to and responsive to the output of each of the first, the second and the third comparing means, for selecting at least a transmit state, wherein the transmit audio path is active, a receive state, wherein the receive audio path is active, or an idle state, wherein neither the transmit or the receive audio paths are active.

Claim 17

A speakerphone as set forth in claim 14 wherein said first generating means, said third generating means, said fourth generating means and said sixth generating means each comprise a signal envelope detection means having a characteristic rise time and a characteristic fall time.

Claim 18

A speakerphone as set forth in claim 15 wherein the rise time of said first generating means and said fourth generating means is approximately one millisecond, wherein the fall time of said first generating means and said fourth generating means is approximately 28 milliseconds, wherein the rise time of said third generating means and said sixth generating means is approximately 0.25 millisecond, and wherein the fall time of said third generating means and said sixth generating means is approximately 100 milliseconds.

Claim 19

A speakerphone as set forth in claim 14 wherein said state selecting means further comprises means for generating a transmit audio path gain defining signal and means for generating a receive audio path gain defining signal and wherein said gain defining signals vary the gain of each of said transmit and said receive audio paths within a range of approximately 0 dB to less than approximately -50 dB when either of said transmit audio path or said receive audio path are active.

Claim 20

A speakerphone as set forth in claim 17 wherein said gain defining signals define, during said idle state, the gain of said transmit audio path to be approximately -10 dB and the gain of said receive audio path to be less than approximately -40 dB.

Claim 23

A method as set forth in claim 19 wherein if the next state of the speakerphone is determined to be other than the present state further includes a step of varying a receive audio path gain signal and a transmit audio path gain signal such that a total loop gain of the transmit and the receive audio paths remains less than unity.

Claim 24

A method as set forth in claim 20 wherein if the next state of the speakerphone is determined to be transmit and the present state is receive or idle the gain of the transmit audio path is varied until it equals approximately zero dB and the gain of the receive audio path is simultaneously varied until it equals less than approximately -50 dB.

Claim 25

A method as set forth in claim 21 wherein if the next state of the speakerphone is determined to be receive and the present state is transmit or idle the gain of the receive audio path is varied until it equals approximately zero dB and the gain of the transmit audio path is simultaneously varied until it equals less than approximately -50 dB.

Claim 26

A method as set forth in claim 20 wherein if the next state of the speakerphone is determined to be idle and the present state is transmit or receive the gain of the receive audio path is varied until it equals less than approximately -40 dB and the gain of the transmit audio path is simultaneously varied until it equals approximately -10 dB.

Claim 27

A method as set forth in claim 22 wherein the gains of the transmit and receive audio paths are varied over an interval of time equal to approximately 20 milliseconds.

Claim 28

A method as set forth in claim 23 wherein the gains of the transmit and receive audio paths are varied over an interval of time equal to approximately 2 seconds.