JPH08288894A - Echo suppression device - Google Patents

Abstract

PURPOSE: To suppress the non-continuous feeling of surrounding noise, to maintain the quality of communication voice and to effectively suppress echo. CONSTITUTION: The transmission side voice signals of a terminal 11 are volume- suppressed in an echo suppression part 12 and sent to the terminal 13. Voice power PTX for one frame of the transmission side voice signals is calculated in a transmission side power calculation part 21 and sent to an echo suppression amount calculation part 23 and the average maximum amplitude SN of the surrounding noise of the transmission side voice signals is sent to the echo suppression part 12. Also, the average PPX(a) of the prescribed frames of reception side voice signals PPX from the terminal 14 is calculated in a reception side power calculation part 24 and sent to the echo suppression amount calculation part 23. In the echo suppression amount calculation part 23, when the average PPX(a) of the voice power on a reception side is equal to or more than a prescribed threshold value TH3 , an echo suppression amount is decided corresponding to the voice power PTX on a transmission side and sent to the echo suppression part 12. In the echo suppression part 12, when the transmission side voice signals are equal to or more than the SN, volume corresponding to the echo suppression amount from the echo suppression amount calculation part 23 is suppressed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an echo suppressor, and more particularly to an echo suppressor for suppressing an echo caused by a sound coming from a speaker into a microphone of a small voice communication terminal such as a mobile phone.

[0002]

2. Description of the Related Art As seen in mobile phones and the like, with the miniaturization of voice communication terminals, voices are circulated from a speaker for receiving into a microphone for transmitting or a voice is coupled to the other party's speaker. The effects of echoes that are perceived on the side cannot be ignored.

In order to remove the echo due to the wraparound of the voice between the transmitting and receiving voices, an echo removing device or an echo that generates a pseudo echo from the receiving voice of the mobile phone and subtracts the echo component from the transmitting voice. A canceller is proposed.

Such an echo removing device has a problem in that the load on the hardware is heavy and that the echo cannot be estimated easily.

As another method, there is also known an echo suppressor or an echo suppressor which uses the volume of the transmitting side and the receiving side of the mobile phone to suppress the volume of the transmitting side when the volume of the receiving side is considered to be high. Has been.

[0006]

By the way, although this echo suppressor requires a relatively small hardware load, it also suppresses the ambient noise mixed in on the transmitting side, so that the receiving end of the other speaker can be received. It may make you feel unnatural.

The present invention has been made in view of the above circumstances, and is a small-sized telephone in which a speaker, which is a sounding means such as a mobile phone, and a microphone, which is a sound collecting means, are arranged close to each other. It is an object of the present invention to provide an echo suppressing device capable of suppressing an unpleasant echo generated on the receiving side of a partner speaker while using a communication terminal and maintaining the quality of a communication voice.

Another object of the present invention is to use an echo canceller which can be realized without using additional hardware by using a part of a DSP (digital signal processor) which is often used in a digital mobile phone or the like. Is to provide.

[0009]

The echo suppressor according to the present invention calculates the ambient noise level and the voice power from the voice signal on the transmitting side from the sound collecting means, and outputs the voice from the voice signal on the receiving side sent to the sound producing means. The power is calculated, the echo suppression amount is calculated based on the voice power of the transmitting side and the voice power of the receiving side, and the voice signal for the transmitting side is calculated based on the echo suppression amount and the ambient noise level. The above problem is solved by performing sound volume suppression.

[0010]

The echo suppression amount of the voice signal on the transmitting side is calculated based on the voice power of the transmitting side and the voice power of the receiving side. The actual volume suppressing operation according to this suppression amount is This is performed only when the voice signal on the transmitting side becomes large with respect to the threshold value based on the ambient noise level of, for example, the average maximum amplitude of ambient noise.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of an echo suppressor according to the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a schematic configuration of an embodiment of an echo suppressor according to the present invention, which is applied to a small communication terminal such as a mobile phone.

In FIG. 1, a transmitting side voice signal from a microphone or the like which is a sound collecting means is input to the terminal 11. Here, the transmitting side voice signal at time n is S [n].
This voice signal S [n] on the transmitting side becomes S ′ [n] via the echo suppressor 12 and is taken out from the output terminal 13. Further, the receiving side voice signal transmitted and received from the communication partner is input to the terminal 14, and is transmitted to the speaker or the like which is the sounding means through the terminal 15. In a small-sized communication terminal such as a mobile phone, a speaker or the like which is a sound producing means and a microphone or the like which is a sound collecting means are arranged close to each other,
By echoing the sound generated by the speaker or the like to the microphone or the like, the echo is recognized on the receiving side of the other speaker.

The transmitting side voice signal S [n] from the terminal 11 is
It is sent to the transmitter power calculator 21 and the ambient noise level calculator 22, and the transmitter power calculator 21 outputs the transmitter voice power P _TX and sends it to the echo suppression amount calculator 23. The average maximum amplitude S _N of the ambient noise on the transmitting side is output from the ambient noise level calculator 22 and sent to the echo suppressor 12. Also, the receiving side voice signal from the terminal 14 is
The average value P _{RX (a)} of the voice power of the receiving side is calculated by being sent to the receiving side power calculating section 24, and this is the echo suppression calculating section 2
Has been sent to 3. In the echo suppression calculation unit 23, _{one of} the echo suppression coefficients γ ₁ and γ ₂ is selected according to a calculation procedure described later, and these echo suppression coefficients γ ₁ and γ are selected.
₂ is sent to the echo suppressor 12.

In such an echo suppressor, a voice is divided into 20 ms units at the time of a call, and is called a frame. At this time, the voice power and the maximum amplitude for one frame on the transmitting side are respectively described above. Voice power P _TX
And the maximum amplitude S _TX , and the voice power for one frame on the receiving side is shown by P _RX .

The ambient noise level calculator 22 calculates the average maximum amplitude S _N of the ambient noise on the transmitting side by the following equation (1).

S _N = (1−α) S _N + αS _TX (0 ≦ α <1) (1) Specifically, for example, at the start of a mobile phone call, the average maximum amplitude S _N is set to an initial value of 0. , The calculation of the above formula (1) is performed a predetermined value N ₁ times for each frame without echo suppression. This resets the count value C ₁ during a call initiation, the count value C ₁ in each frame increases i.e. incremented by one, until the count value C ₁ reaches the predetermined value N _1, the above equation (1) Should be calculated.

However, when it is considered that voice exists on the transmitting side, for example, when the voice power P _TX on the transmitting side is larger than a predetermined threshold value TH ₁ (P _TX > TH ₁ ), the above equation (1) is used. The count value C ₁ is also fixed without performing the calculation.

In the receiver side power calculator 24, the count value C
_{After 1} exceeds the predetermined value N ₁ , the voice power P _RX of the receiving side is obtained by tracing back to the past N ₂ frames, and the average value P _{RX (a)} of the voice power P _RX for these N ₂ frames is obtained. To calculate.

Here, as a specific example of each of the predetermined values N ₁ and N ₂ , when one frame is 20 ms as described above,
N ₁ is about 50 (frames) corresponding to about 1 second, and N
₂ may be about 5 (frames) corresponding to about 100 ms.

When the average value P _{RX (a) of the} voice power on the receiving side becomes equal to or more than the threshold value TH _{3 which} is considered to be mixed as an echo in the microphone, the echo suppression amount calculation unit 23
That is, when P _{RX (a)} ≧ TH ₃ , the volume on the transmitting side is suppressed under the following conditions based on the voice power P _TX on the transmitting side and the like. At this time, the above formula (1) is not applied, and when the count value C ₁ is smaller than the predetermined value N ₁ , it is incremented by 1.

Here, TH ₂ is a threshold value of the voice power which is considered to be the presence of some voice on the transmitting side in addition to the echo.

The echo suppressing operation is performed in accordance with the relationship between the voice power P _TX on the transmitting side and the thresholds TH ₁ and TH ₂ .
For example, switching control is performed to three states a, b, and c as shown in FIG.

First, the state a is the voice power P on the transmitting side.
_The above-mentioned threshold T at which _TX is considered to have voice on the transmitting side
This is the case when the volume is high such that H ₁ or more (P _TX ≧ TH ₁ ). In this case, volume suppression is not performed, and the output signal S '[n] output from the echo suppressor 12 is equal to the voice signal S [n] on the transmitting side (S' [n] = S [n]). Become.

Next, the state b is the voice power P on the transmitting side.
_This is a case of medium sound volume in which _TX is smaller than the threshold TH ₁ and is equal to or larger than the threshold TH ₂ (TH ₂ ≦ P _TX <TH ₁ ). In this case, the echo suppression coefficient γ ₁ (however,
The echo suppression operation by 0 ≦ γ ₁ <1) is performed, and specifically, the voice signal S [n] on the transmitting side is smaller than the average maximum amplitude S _N of the ambient noise on the transmitting side or S _N. Depending on the above, the volume suppression operation is switched on and off. That is, when S [n] <S _N , volume suppression is not performed and S
Only when [n] ≧ S _N , such as S ′ [n] = sign (S [n]) × (S _N ＋ (abs (S [n]) − S _N ) × γ ₁ ) (2) Volume suppression is performed.

Next, the state c is the voice power P on the transmitting side.
_This is a case where the volume of _TX is small such that _TX is smaller than the threshold TH ₂ (P _TX <TH ₂ ). In this case, the echo suppression operation by the echo suppression coefficient γ ₂ (where 0 ≦ γ ₂ <γ ₁ ) is performed, and specifically, the transmitting side voice signal S [n] is
The sound volume suppressing operation is switched on and off depending on whether it is smaller than the average maximum amplitude S _N of the ambient noise on the transmitting side or above S _N. That is, when S [n] <S _N , volume suppression is not performed, and only when S [n] ≧ S _N , S ′ [n] = sign (S [n]) × (S _N + ( Volume suppression such as abs (S [n])-S _N ) × γ ₂ ) (3) is performed.

Here, each part in the configuration shown in FIG.
It may be implemented by software using a so-called DSP (digital signal processor), or may be configured by hardware.

Next, FIG. 3 is a flow chart for explaining the above-mentioned echo suppressing operation.

Prior to the first call of the echo suppressing operation of FIG. 3, the state flag state is set to "0".

In the first step S1 of FIG. 3, the value of the state flag state is confirmed. If it is "0", the process proceeds to step S2, and the average maximum amplitude S _N of the ambient noise on the transmitting side is confirmed. Also, the count value C ₁ is cleared or reset to 0 and the state flag state is set to “1”.
And set to step S3. When it is determined in step S1 that the state flag state is not "0", the process directly proceeds to step S3.

In step S3, it is determined whether or not the voice power P _TX on the transmitting side is equal to or more than the threshold value TH ₁ (P _TX ≧ TH ₁ ), and if Yes, the state a is set and the sound volume is not suppressed. The process ends. If No, go to the next step S
In step 4, it is determined whether the average value P _{RX (a)} of the voice power on the receiving side is equal to or more than the threshold value TH ₃ (P _{RX (a)} ≧ TH ₃ ).

[0032] In the step S4, so as not to affect the echo in the calculation of average maximum amplitude S _N of the ambient noise of the transmitting side, when the No only, by the following step S5 to calculate the S _N S After updating _N , step S6
If Yes, the process proceeds to step S6 without updating S _N.

In step S6, it is determined whether or not the count value C ₁ has reached the predetermined value N ₁ , and if No, the count value C ₁ is incremented in step S7, that is, C ₁ = C ₁ +1. After that, the process ends. If Yes is determined in step S6, the process proceeds to step S8.

In step S8, the average value P _{RX (a)} of the voice power on the receiving side is not less than the threshold value TH ₃ (P _{RX (a)} ≧ T
H ₃ ), and if No, terminates the process and returns Y
If es, the process proceeds to step S9, and the above-mentioned state b
Echo suppression processing corresponding to the state c is performed.

According to the embodiment of the present invention as described above, it is possible to suppress the unpleasant echo that occurs on the receiving side of the other speaker when using the mobile phone or the like, and to maintain the quality of the communication voice. Become. Moreover, it is possible to suppress the discontinuity of ambient noise found in a conventional simple echo suppressor and effectively perform echo suppression. Also, a part of a DSP (digital signal processor) that is often used in a digital mobile phone or the like is used, and it is possible to mount the digital mobile phone in a relatively small modification without requiring additional hardware. You can

[0036] The present invention is not limited to the embodiments described above, for example, an echo suppression coefficient and gamma _1, gamma ₂ of two types in the embodiment, the transmission side of the speech power P _TX Depending on the result of comparison with the threshold value TH ₂ , these coefficients γ
_{Although 1} and γ ₂ are switched, the echo suppression coefficient may be one type or three or more types. Of course, various modifications can be made without departing from the scope of the present invention.

[0037]

According to the echo suppressor of the present invention,
The ambient noise level and voice power are calculated from the voice signal on the transmitting side from the sound collecting means, the voice power is calculated from the voice signal on the receiving side sent to the sound generating means, and the voice power on the transmitting side and the voice power on the receiving side are calculated. The echo suppression amount is calculated based on the above, and the volume of the voice signal on the transmitting side is suppressed based on the echo suppression amount and the ambient noise level. It is possible to suppress an unpleasant echo that occurs on the receiving side of.

Furthermore, when the level of the voice signal on the transmitting side is smaller than the average maximum amplitude of the ambient noise, the volume is not suppressed, and only when the level of the average maximum amplitude of the ambient noise is larger than that, the volume of the voice signal on the transmitting side is suppressed. By doing so, it is possible to suppress the discontinuity of ambient noise and maintain the quality of the communication voice.

[Brief description of drawings]

FIG. 1 is a block diagram schematically showing a basic configuration of an echo suppressor according to an embodiment of the present invention.

FIG. 2 is a diagram showing input / output characteristics of an echo suppressor according to an embodiment of the present invention.

FIG. 3 is a flowchart for explaining the operation of the echo suppressor according to the embodiment of the present invention.

[Explanation of symbols]

 12 Echo Suppressor 21 Transmitter-side Power Calculator 22 Ambient Noise Level Calculator 23 Echo Suppression Amount Calculator 24 Receiver-side Power Calculator

Claims (2)

[Claims] 1. An echo suppressor for suppressing an echo generated when a sound from a sounding means wraps around a sound collecting means arranged in proximity to the sound collecting means. Ambient noise level calculating means, a transmitting side power calculating means for calculating voice power from the transmitting side voice signal, and a receiving side power calculation for calculating voice power from a receiving side voice signal sent to the sounding means. Means, an echo suppression amount calculation means for calculating an echo suppression amount based on the voice power from the transmitting side power calculating means, and the voice power from the receiving side power calculating means, and the echo suppression amount calculating means. Echo suppression means for suppressing the volume of the voice signal on the transmitting side based on the output from the above and the output from the ambient noise level calculation means. Location. 2. The ambient noise level calculating means calculates an average maximum amplitude of ambient noise of the transmitting side voice signal, and the echo suppressing means calculates the average of the ambient noise of the transmitting side voice signal. The echo suppressor according to claim 1, wherein the volume is not suppressed when the amplitude is smaller than the maximum amplitude, and the volume of the voice signal on the transmitting side is suppressed only when the average maximum amplitude of the ambient noise is larger than the maximum amplitude.