CN1379551A - Echo eliminator - Google Patents

Echo eliminator Download PDF

Info

Publication number
CN1379551A
CN1379551A CN 01110298 CN01110298A CN1379551A CN 1379551 A CN1379551 A CN 1379551A CN 01110298 CN01110298 CN 01110298 CN 01110298 A CN01110298 A CN 01110298A CN 1379551 A CN1379551 A CN 1379551A
Authority
CN
China
Prior art keywords
signal
convergence
echo
constant
input
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN 01110298
Other languages
Chinese (zh)
Other versions
CN1172451C (en
Inventor
陈宗奎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Winbond Electronics Corp
Original Assignee
Winbond Electronics Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Winbond Electronics Corp filed Critical Winbond Electronics Corp
Priority to CNB011102985A priority Critical patent/CN1172451C/en
Publication of CN1379551A publication Critical patent/CN1379551A/en
Application granted granted Critical
Publication of CN1172451C publication Critical patent/CN1172451C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)

Abstract

An echo suppressor between an input line and an output line contains an adjustable filter, an echo-adjustment judgement device, a double-conversation detector, and a newly added divergence detector, among which the adjustable filter suppresses echo after adjusting its adjustment coefficient, and the echo-adjustment judgement device outputs an adjustment judgement signal according to the acoustic input and output signals, and the divergence detector sends out a divergence-convergence signal after receiving the outside input signal and the adjustment judgement signals, the double-conversation detector sends out the control adjustment coefficient signal to the adjustable filter after receiving the outside input signal, the adjustment judgement signal and the divergence-convergence signal.

Description

Echo canceller
The present invention relates to a kind of echo canceller, particularly relate to a kind of echo canceller of dispersing detector that has, disperse detector by this and coefficient situation of dispersing of echo canceller is improved be tending towards convergence, thereby make echo canceller can stablize running.
Echo canceller is usually used in the phone, and in order to eliminate unnecessary echo, for example 2 lines change 4 lines in the phone networking, the echo that does not match and produced because of impedance, or when using speaker-phone, because at confined space, sound can reflect and cause echo.And echo canceller can only be eliminated unidirectional echo, and can only adjust its coefficient when unidirectional (producing the direction of echo) conversation, coefficient can not be adjusted under other situations, otherwise can cause the not good situation of coefficient convergence, so must survey one-way traffic with dual dialog detection device.Dual dialog detection device has detection method miscellaneous, and common has following two kinds:
Figure 1 shows that the structure chart of known echo canceller, wherein echo canceller 14 is connected between a reception line 10 and the output line 12, receive the first input end 16 of an end of line 10 as external input signal Rin, the other end is as first output 18 of external output signal Rout, one end of output line 12 is as second input 20 of audio input signal Sin, and the other end is as second output 22 of voice output signal Sout.Wherein external output signal Rout sends through loud speaker 24, and microphone 26 is then sent audio input signal Sin at second input 20.Between loud speaker 24 and microphone 26, because the signal that loud speaker 24 is sent can be overlapped in the audio input signal Sin that microphone 26 is sent, thereby cause producing echo 28, at this moment the dual dialog detection device in the echo canceller just must be surveyed and adjust echo canceller, and above-mentioned echo 28 is effectively eliminated.
With the dual dialog detection device of first kind of echo canceller,, promptly be judged as dual dialogue, and freeze the adjustment of echo canceller coefficient at once when the size of audio input signal Sin surpasses the half of external output signal Rout.This method can be used for linear echo canceller, but is not useable for the needed sense of hearing echo canceller of loud speaker, far has closely because the speaker has from the distance of microphone, and the volume of echo also may vary, and is unpredictable fully, so be difficult to satisfy above-mentioned condition.
With the dual dialog detection device of second kind of type echo canceller, it uses ERLE (echo loss enhancing) to judge.
ERLE=sin/sout
When ERLE just can remove to adjust the coefficient of echo canceller during greater than certain constant.This is after the coefficient convergence of echo canceller, and ERLE is effective judgment mode really.If when coefficient is not restrained as yet or dispersed, thereby ERLE may cause erroneous judgement to pin the coefficient of echo canceller, makes echo canceller can't restrain forever.
Therefore, the objective of the invention is to design one and disperse detector and untie this locking-in state, make that echo canceller can stable again running.
In addition, the present invention also aims to provide a kind of above-mentioned method that the echo canceller of dispersing detector carries out echo cancellation that has of utilizing.
For reaching above-mentioned and other purposes, according to a kind of echo canceller provided by the invention, be applicable to that one receives between a line and the output line, wherein should receive the first input end of an end of line as an external input signal, the other end is as one first output of an external output signal, one end of this output line is as one second input of an audio input signal, and the other end is as one second output of a voice output signal, and this echo canceller comprises:
One adjusting type filter is connected between first output and second input, and it is in order to receiving this external output signal, and produces an echo cancellation signal to carry out echo cancellation after to the adjustment of adjusting type filter coefficient; Determining device is adjusted in one echo, is connected to second input and second output, in order to export an adjustment judgement signal according to audio input signal and voice output signal; One subtracter is connected to second input, second output and adjusting type filter, in order to output sound output signal after receiving audio input signal and echo cancellation signal; One disperses detector, is connected to first input end and echo and adjusts determining device, judges in order to disperse convergence after receiving external input signal and adjusting the judgement signal, and sends one with this and disperse the convergence signal; An and dual dialog detection device, be connected to first input end, adjusting type filter, echo adjustment determining device and disperse detector, in order to after receiving external input signal, adjust the judgement signal and dispersing the convergence signal, send a control and adjust coefficient signal to the adjusting type filter; Wherein, described second output is sent one and is adjusted the echo cancellation signal that the echo cancellation signal is controlled the adjusting type filter.
Aforesaid echo canceller, wherein said voice output signal can deduct the echo cancellation signal for audio input signal.
In addition, the adjustment of above-mentioned echo adjustment determining device output judges that signal is the ratio of audio input signal divided by the voice output signal.And dual dialog detection device adjust to be judged signal less than a convergence constant, or the absolute value of external input signal is during less than an input constant, and then control is adjusted coefficient signal the adjusting type filter coefficient is freezed, and wherein convergence constant is greater than one.
In addition, for dispersing detector, when the absolute value of external input signal greater than the input constant and adjust when judging signal greater than convergence constant, the convergence signal of sending of dispersing is a convergence signal, if when the absolute value of external input signal is dispersed constant greater than the adjustment judgement signal of input constant and reception less than one, then send disperse the convergence signal be to disperse signal, this disperses signal makes control that dual dialog detection device sent adjust coefficient signal the adjusting type filter coefficient to be removed, wherein dispersed constant less than one.
Also a kind of echo removing method according to the present invention, be applicable between an external input signal, an external output signal, an audio input signal and the voice output signal, this echo removing method comprises: produce an echo cancellation signal, pass to the influence that audio input signal produces in order to eliminate external output signal; According to described audio input signal and described voice output signal, produce one and adjust the judgement signal; Judge signal according to external input signal and described adjustment, produce one and disperse the convergence signal; Judge signal and the described convergence signal of dispersing according to external input signal, described adjustment, produce a control adjustment coefficient signal and adjust described echo cancellation signal in order to control; And according to audio input signal and echo cancellation signal and the output sound output signal, and send one and adjust the echo cancellation signal in order to control described echo cancellation signal.
In addition, above-mentioned adjustment judges that signal can be the ratio of audio input signal divided by the voice output signal.When signal was judged in described adjustment less than a convergence constant, then described control was adjusted coefficient signal to the echo cancellation signal-freeze, and wherein this convergence constant is greater than one.When the absolute value of external input signal was imported constant less than one, then described control was adjusted coefficient signal to the echo cancellation signal-freeze.
In addition, when the absolute value of external input signal greater than the input constant and adjust when judging signal greater than convergence constant, the convergence signal of sending of dispersing is a convergence signal, if when the absolute value of external input signal is dispersed constant greater than the adjustment judgement signal of described input constant and reception less than one, then send disperse the convergence signal be to disperse signal, this disperses signal makes control adjustment coefficient signal remove the echo cancellation signal, and wherein this disperses constant less than one.
Compare with known echo canceller, echo canceller of the present invention has increased disperses detector, receive external input signal and adjust by it and judge that signal is controlled and send one and disperse the dual dialog detection device of convergence signal, and then effectively control the variation of adjusting type filter coefficient, make that originally dispersing the situation that can't restrain obtains changing.Method proposed by the invention is dispersed the convergence signal by providing, is adjusted and judge signal, makes the echo cancellation signal effectively adjust, and will originally disperse to be improved to be tending towards restraining situation.Thereby obtain better echo cancellation effect.
For making above-mentioned and other purposes of the present invention, feature and advantage clearer understandable, the present invention is described in more detail below in conjunction with accompanying drawing.
Fig. 1 is the structure chart of known echo canceller; And
Fig. 2 is the structure chart according to the echo canceller of one embodiment of the invention.
With reference to figure 2, it is depicted as the structure chart according to a kind of echo canceller of one embodiment of the invention.Our echo canceller of the present invention as can be seen comprises adjusting type filter 50, echo adjustment determining device 52, subtracter 65, disperses detector 54 and dual dialog detection device 56 from figure.
Echo canceller of the present invention is connected one and receives between a line 58 and the output line 60, and an end of wherein above-mentioned reception line 58 is as the first input end 62 of an external input signal Rin, and the other end is as first output 64 of an external output signal Rout.And an end of output line 60 is as second input 66 of audio input signal Sin, and the other end is as second output 68 of voice output signal Sout.
Adjusting type filter 50 among the present invention is connected between first output 64 (node 63 expressions of extending with first output 64 among the figure) and second input 66, it act as and receives external output signal Rout, and produces an echo cancellation signal 70 after to the coefficient adjustment of inner adjusting type filter and carry out echo cancellation.
Echo adjustment determining device 52 (representing with ERLE among the figure) is connected to second input 66 (representing with the node 67 of second input, 66 extensions among the figure) and second output 68 (node 69 expressions of extending with second output 68 among the figure), in order to export an adjustment judgement signal 72 according to audio input signal Sin and voice output signal Sout.Wherein, the adjustment judgement signal 72 that determining device 52 is exported is adjusted in echo, for example uses the ratio of audio input signal Sin divided by voice output signal Sout, and its mathematical expression is as follows:
ERLE=sin/sout
Subtracter 65 is connected to second input 66, second output 68 and adjusting type filter 50, in order to receive audio input signal Sin and echo cancellation signal 70 back output sound output signal Sout.
Be connected to first input end 62 (node 61 expressions of extending with first input end 62 among the figure) and adjust determining device 52 as for dispersing 54 of detectors, in order to disperse convergence after judging signal 72 and judge and send one and disperse convergence signal 74 (below represent) with adjusting receiving external input signal Rin with D with echo.
Dual dialog detection device 56 is connected to first input end 62 (node 61 expressions of extending with first input end 62 among the figure), adjusting type filter 50, echo and adjusts determining device 52 and disperse detector 54, in order to receive external input signal Rin, to adjust and send a control after judging signal 72 and dispersing convergence signal 74 and adjust coefficient signal 76 to adjusting type filter 50.
Wherein, adder 65 is depicted as audio input signal Sin and deducts echo cancellation signal 70 among voice output signal Sout such as the figure, and second output 68 (node 77 expressions of extending with second output 68 among the figure) is sent the echo cancellation signal 70 that an adjustment echo cancellation signal 78 is controlled adjusting type filter 50.
Under general situation, dual dialog detection device 56 externally input signal Rin value absolute value less than input constant K r (promptly | during Rin|<Kr), or adjust when judging signal 72 (ERLE) less than a convergence constant Kc, then 76 pairs of adjusting type filter 50 coefficients of control adjustment coefficient signal freeze.Wherein above-mentioned convergence constant Kc is greater than one (Kc>1).
Disperse detector 54 when the absolute value of external input signal Rin greater than input constant K r (promptly | Rin|>Kr) and adjusting judges that (during ERLE>Kc), the convergence signal of then sending 74 (D) of dispersing is one to restrain signal (for example making D=0) to signal 72 (ERLE) greater than convergence constant Kc.If when the absolute value of external input signal Rin greater than input constant K r (promptly | Rin|>Kr) and the adjustment of reception judgement signal 72 (ERLE) are dispersed constant K d (during ERLE<Kd) less than one, then send disperse the convergence signal 74 be to disperse signal (for example making D=1), disperse signal and make dual dialog detection device 56 send control to adjust 76 pairs of adjusting type filter coefficients of coefficient signal and remove, the wherein above-mentioned constant K d that disperses is less than one (Kd<1).
Disperse convergence signal 74 when dispersing signal (D=1) when what send, represent that then the adjusting type filter 50 of echo canceller dispersed, this situation generally can occur in the echo environment to be had when significantly improving, and for example there are objects moving or the open and close door in loud speaker uses.Must revise original dual dialog detection device 56 this moment, when disperse convergence signal 74 for the absolute value of dispersing signal (D=1) and external input signal Rin greater than input constant K r (promptly | during Rin|<Kr), perhaps when disperse convergence signal 74 for the absolute value of convergence signal (D=0) and original external input signal Rin value less than import constant K r (promptly | during Rin|<Kr), or adjust when judging signal 72 (ERLE) less than a convergence constant Kc, then 76 pairs of adjusting type filter 50 coefficients of control adjustment coefficient signal freeze.
When dispersing convergence signal 74 when dispersing signal (D=1), adjust to judge that then signal 72 (ERLE) do not have a reference value, can only rely on the absolute value of external input signal Rin value | whether Rin| decides adjusting type filter 50 coefficients to adjust, thereby make coefficient to restrain gradually, cause and disperse convergence signal 74 for restraining signal (D=0), dual dialog detection device 56 returns to normal general state again, thereby makes echo canceller of the present invention reach stable.
In sum, echo canceller of the present invention is being dispersed the down auxiliary of detector, makes original divergent cases be tending towards convergence, and dual dialog detection device can be judged with normal condition, thereby makes the stable running of echo canceller.
Though invention has been described with preferred embodiment above; yet it is not in order to limit the present invention; any this field skilled person; without departing from the spirit and scope of the present invention; can make various distortion and modification, so protection scope of the present invention should be with being as the criterion that claims were defined.

Claims (17)

1. echo canceller, be applicable to that one receives between a line and the output line, wherein should receive the first input end of an end of line as an external input signal, the other end is as one first output of an external output signal, one end of this output line is as one second input of an audio input signal, the other end is as one second output of a voice output signal, and it is characterized in that: this echo canceller comprises:
One adjusting type filter is connected between first output and second input, and it is in order to receiving this external output signal, and produces an echo cancellation signal to carry out echo cancellation after to the adjustment of adjusting type filter coefficient;
Determining device is adjusted in one echo, is connected to second input and second output, in order to export an adjustment judgement signal according to audio input signal and voice output signal;
One subtracter is connected to second input, second output and adjusting type filter, in order to output sound output signal after receiving audio input signal and echo cancellation signal;
One disperses detector, is connected to first input end and echo and adjusts determining device, judges in order to disperse convergence after receiving external input signal and adjusting the judgement signal, and sends one with this and disperse the convergence signal; And
One dual dialog detection device, be connected to first input end, adjusting type filter, echo and adjust determining device and disperse detector, in order to receive external input signal, adjust judge signal and disperse the convergence signal after and send a control and adjust coefficient signal to the adjusting type filter;
Wherein, described second output is sent one and is adjusted the echo cancellation signal that the echo cancellation signal is controlled the adjusting type filter.
2. echo canceller as claimed in claim 1 is characterized in that: described echo is adjusted the adjustment of determining device output and is judged that signal is the ratio of audio input signal divided by the voice output signal.
3. echo canceller as claimed in claim 2, it is characterized in that: described dual dialog detection device, when adjusting the judgement signal less than a convergence constant, then described control is adjusted coefficient signal the adjusting type filter coefficient is freezed, and wherein said convergence constant is greater than one.
4. echo canceller as claimed in claim 2 is characterized in that: described dual dialog detection device, when externally the absolute value of input signal was less than an input constant, then described control was adjusted coefficient signal the adjusting type filter coefficient is freezed.
5. echo canceller as claimed in claim 2, it is characterized in that: the described detector of dispersing, when the absolute value of external input signal greater than the input constant and adjust when judging signal greater than convergence constant, the convergence signal of sending of dispersing is a convergence signal, if when the absolute value of external input signal is dispersed constant greater than the adjustment judgement signal of input constant and reception less than one, then send disperse the convergence signal be to disperse signal, this disperses signal makes control that dual dialog detection device sent adjust coefficient signal the adjusting type filter coefficient is removed, and the wherein said constant of dispersing is less than one.
6. echo canceller as claimed in claim 5 is characterized in that: when dispersing the convergence signal is the absolute value during greater than the input constant of dispersing signal and external input signal, and then described control is adjusted coefficient signal the adjusting type filter coefficient is freezed.
7. echo canceller as claimed in claim 5 is characterized in that: when dispersing the convergence signal for the absolute value of convergence signal and external input signal during less than the input constant, then described control is adjusted coefficient signal the adjusting type filter coefficient is freezed.
8. echo canceller as claimed in claim 5 is characterized in that: when dispersing the convergence signal for the convergence signal and when adjust judging that signal is less than a convergence constant, then described control is adjusted coefficient signal the adjusting type filter coefficient is freezed.
9. echo canceller as claimed in claim 1 is characterized in that: the voice output signal that described subtracter is exported is that audio input signal deducts the echo cancellation signal.
10. echo removing method is applicable to that it is characterized in that: described echo removing method comprises between an external input signal, an external output signal, an audio input signal and the voice output signal:
Produce an echo cancellation signal, pass to the influence that audio input signal produces in order to eliminate external output signal;
According to described audio input signal and described voice output signal, produce one and adjust the judgement signal;
Judge signal according to external input signal and described adjustment, produce one and disperse the convergence signal;
Judge signal and the described convergence signal of dispersing according to external input signal, described adjustment, produce a control adjustment coefficient signal and adjust described echo cancellation signal in order to control; And
According to audio input signal and echo cancellation signal and the output sound output signal, and send one and adjust the echo cancellation signal in order to control described echo cancellation signal.
11. echo removing method as claimed in claim 10 is characterized in that: described adjustment judges that signal is the ratio of audio input signal divided by the voice output signal.
12. echo removing method as claimed in claim 11 is characterized in that: when adjusting the judgement signal less than a convergence constant, then described control is adjusted coefficient signal to the echo cancellation signal-freeze, and wherein this convergence constant is greater than one.
13. echo removing method as claimed in claim 11 is characterized in that: when the absolute value of external input signal was imported constant less than one, then described control was adjusted coefficient signal to the echo cancellation signal-freeze.
14. echo removing method as claimed in claim 13, it is characterized in that: when the absolute value of external input signal greater than the input constant and adjust when judging signal greater than convergence constant, the convergence signal of sending of dispersing is a convergence signal, if when the absolute value of external input signal is dispersed constant greater than the adjustment judgement signal of described input constant and reception less than one, then send disperse the convergence signal be to disperse signal, this disperses signal makes control adjustment coefficient signal remove the echo cancellation signal, and wherein this disperses constant less than one.
15. echo removing method as claimed in claim 13 is characterized in that: when dispersing the convergence signal is the absolute value during greater than the input constant of dispersing signal and external input signal, and then described control is adjusted coefficient signal to the echo cancellation signal-freeze.
16. echo removing method as claimed in claim 13 is characterized in that: when dispersing the convergence signal for the absolute value of convergence signal and external input signal during less than the input constant, then described control is adjusted coefficient signal to the echo cancellation signal-freeze.
17. echo removing method as claimed in claim 13 is characterized in that: when dispersing the convergence signal for the convergence signal and when adjust judging that signal is less than a convergence constant, then described control is adjusted coefficient signal to the echo cancellation signal-freeze.
CNB011102985A 2001-04-06 2001-04-06 Echo eliminator Expired - Fee Related CN1172451C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB011102985A CN1172451C (en) 2001-04-06 2001-04-06 Echo eliminator

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNB011102985A CN1172451C (en) 2001-04-06 2001-04-06 Echo eliminator

Publications (2)

Publication Number Publication Date
CN1379551A true CN1379551A (en) 2002-11-13
CN1172451C CN1172451C (en) 2004-10-20

Family

ID=4658488

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB011102985A Expired - Fee Related CN1172451C (en) 2001-04-06 2001-04-06 Echo eliminator

Country Status (1)

Country Link
CN (1) CN1172451C (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100344074C (en) * 2004-08-16 2007-10-17 何顺兰 Plate card type echo eliminator
CN101119135B (en) * 2007-07-04 2010-09-01 深圳市融创天下科技发展有限公司 Step parameter regulation means and equipment for eliminating echo
CN102340605A (en) * 2011-07-25 2012-02-01 华为技术有限公司 Method and device for cancelling echoes

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100344074C (en) * 2004-08-16 2007-10-17 何顺兰 Plate card type echo eliminator
CN101119135B (en) * 2007-07-04 2010-09-01 深圳市融创天下科技发展有限公司 Step parameter regulation means and equipment for eliminating echo
CN102340605A (en) * 2011-07-25 2012-02-01 华为技术有限公司 Method and device for cancelling echoes
CN102340605B (en) * 2011-07-25 2013-06-05 华为技术有限公司 Method and device for cancelling echoes

Also Published As

Publication number Publication date
CN1172451C (en) 2004-10-20

Similar Documents

Publication Publication Date Title
JP3727258B2 (en) Echo suppression processing system
US7856097B2 (en) Echo canceling apparatus, telephone set using the same, and echo canceling method
EP1832104B1 (en) Audio system and method for acoustic echo cancellation
US20100208908A1 (en) Echo supressing method and apparatus
EP1022866A1 (en) Echo elimination method, echo canceler and voice switch
US6459942B1 (en) Acoustic coupling compensation for a speakerphone of a system
WO2003044978A1 (en) Echo canceller ensuring further reduction in residual echo
CN111052767B (en) Audio processing device, audio processing method, and information processing device
CN1533665A (en) Echo canceller having nonlinear echo suppressor for harmonics calculations
JP2007312364A (en) Equalization in acoustic signal processing
CN1383314A (en) Method for modulating noise shielding and noise interfrence in speech communication
EP1700465A1 (en) System and method for enchanced subjective stereo audio
KR100633213B1 (en) Methods and apparatus for improving adaptive filter performance by inclusion of inaudible information
US5790657A (en) Echo suppressor capable of suppressing an echo resulting from acoustic coupling without spoiling a natural sound of conversation
JPH10257159A (en) Loud-speaker communication equipment
CN1172451C (en) Echo eliminator
JP2003051879A (en) Speech device
CN1322427A (en) Methods and apparatus for mitigating efects of microphone overload in echo cancelation systems
KR101185650B1 (en) Method and apparatus for eliminating acoustic echo from voice signal
JP3220979B2 (en) Voice switch
KR100431965B1 (en) Apparatus and method for removing echo-audio signal using time-varying algorithm with time-varying step size
KR100790042B1 (en) Noise canceling device and method thereof, and communication terminal using the same
JP2843782B2 (en) Loudspeaker
CN117690443A (en) Voice processing method and device, electronic equipment and storage medium
JPS63294019A (en) Howling suppression device

Legal Events

Date Code Title Description
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C17 Cessation of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20041020