JPH08223275A - Hand-free talking device - Google Patents

Abstract

PURPOSE: To effectively suppress howling and to realize a hand-free talking device capable of two-way simultaneous talking inexpensively. CONSTITUTION: The device is made up of a speaker 30, plural microphones 31a, 31b arranged in the same case 3 with a different distance from the speaker 30, transmission signal amplifier means 21a, 21b provided corresponding to the microphones 31a, 31b and amplifying an output signal from the microphones 31a, 31b respectively, a reception signal amplifier means 20 amplifying an input signal to the speaker 30, a direct coupling suppression processing section 50 suppressing a direct coupling component of a reception signal included in a transmission signal by means of a filter 502 and an arithmetic processing means 503 and a voice switch 1 suppressing an acoustic coupling component of the reception signal included in the transmission signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication device used for intercoms, telephones and the like, and more particularly to a hands-free voice communication device having a function of preventing howling.

[0002]

2. Description of the Related Art In a communication device equipped with a speaker and a microphone, when the call volume is raised, a part of the audio signal emitted from the speaker is fed back to the microphone, causing an oscillation phenomenon called howling. In particular,
Hands-free communication devices such as intercoms and telephones
There is a problem that howling is likely to occur because the distance between the speaker for reception and the microphone for transmission is as short as several cm to several tens of cm compared to the distance with the caller.

In order to prevent the occurrence of such howling, the hands-free speech communication apparatus needs to perform acoustic coupling suppression processing, and the main acoustic coupling suppression processing methods conventionally used in hands-free speech communication apparatuses are as follows. There are a method using a voice switch and a method using a voice switch and an echo canceller together. However, when a method using a voice switch and an echo canceller together is adopted, there is a problem that the cost becomes high.

FIG. 15 shows a schematic structure of the voice switch. This method compares the levels of a received signal and a transmitted signal and changes the loss amount of a variable loss circuit provided for attenuating the received signal or the transmitted signal based on the comparison result. That is, in this system, a loss is inserted in a closed loop formed by acoustic coupling from a speaker to a microphone to reduce the closed loop loop gain, and a howling margin and an echo return loss are secured.

FIG. 16 shows the relationship between the level difference between the reception signal and the transmission signal of this voice switch and the insertion loss for attenuating the respective signals. In the transmission mode, the insertion loss into the reception signal is large and the insertion loss into the transmission signal is small, while in the reception mode, the insertion loss into the reception signal is small and the insertion loss into the transmission signal is small. Is getting bigger.

[0006]

However, in order to obtain the howling margin and the echo return loss required for a comfortable call by the above method using the voice switch, the required insertion loss amount is required. There was a problem that it became very large and a feeling of disconnection was felt during a call.

When the acoustic coupling component contained in the transmission signal is larger than the reception signal, the voice switch is switched to the transmission mode and the reception signal is attenuated,
A phenomenon called so-called reception blocking occurs in which the required reception volume cannot be obtained and the user cannot hear the volume. As described above, this method has a problem in that it is not possible to realize a two-way simultaneous call because there is a loss in one of the call paths.

The present invention has been made in view of the above circumstances, and provides a hands-free communication device that can effectively suppress the occurrence of howling and enable two-way simultaneous communication at a relatively low cost. The purpose is to do.

[0009]

A hands-free communication device according to the present invention according to claim 1 is a speaker for converting a received signal into a voice and outputting the voice, and a speaker for changing a distance from the speaker. Placed in the same cabinet,
Two microphones for converting the transmitted voice into an electric signal and outputting the electric signal, that is, a first microphone arranged in a position close to the speaker, a second microphone arranged in a position far from the speaker, and a second microphone First transmission signal amplification means for amplifying the output signal of the one microphone, second transmission signal amplification means for amplifying the output signal of the second microphone, and reception signal for amplifying the input signal to the speaker. A filter for performing a convolution operation on the amplification means and a transfer function representing the ratio of the output signal of the first transmission signal amplification means and the transfer function of the acoustic direct propagation system directly reaching a microphone from a speaker identified in advance; An acoustic direct coupling suppression processing unit including arithmetic processing means for obtaining a difference between the output signal of the filter and the output signal of the second transmission signal amplifying means, and the transmission signal Constituted by a voice switch that suppresses acoustic coupling component of the received signal contained.

According to the second aspect of the present invention, the characteristics of the ratio of the outputs of the two microphones are identified in the direct coupling suppression processing section of the handsfree speech device according to the first aspect. , An adaptive filter that performs a convolution operation of the output signal of the first transmission signal amplification means and the identified transfer function, and the difference between the output signal of the adaptive filter and the output signal of the second transmission signal amplification means. And an oscillation determination means for determining the possibility of oscillation based on the amplitude ratio between the output signal of the arithmetic processing means and the input signal of the reception signal amplifier and outputting the determination result to the voice switch. And are provided.

According to the third aspect of the present invention, the hands-free speech communication apparatus according to the present invention corrects the distortion generated in the transmission signal in the output of the direct coupling suppression processing section of the hands-free speech communication apparatus according to the first or second aspect. And a distortion correction means for the purpose. According to the fourth aspect of the present invention, the hands-free communication device according to the present invention is arranged in the same casing as the speaker that converts the received signal into voice and outputs the voice and the speaker so that the distance from the speaker is different. To three or more microphones for converting the electric signals into electric signals and outputting the electric signals, and three or more transmission signal amplifying means provided corresponding to the respective microphones for amplifying the output signals of the respective microphones, and to the speaker. A reception signal amplifying means for amplifying an input signal, an acoustic direct coupling suppression processing section having an arithmetic processing means for calculating a sum of output signals of the transmission signal amplifying means, and an acoustic coupling component of a reception signal included in the transmission signal. And a voice switch that suppresses the noise.

According to a fifth aspect of the present invention, there is provided a hands-free speech communication apparatus, wherein the output of the direct coupling suppression processing unit of the hands-free speech communication apparatus according to any one of the first to fourth aspects transmits the transmitted transmission signal. Predetermined as a band to be
It is configured by providing a filter that selectively passes only predetermined frequency components. A hands-free communication device according to the present invention described in claim 6 is a band to be transmitted of the output transmission signal to the output of the transmission signal amplification means of the hands-free communication device according to any one of claims 1 to 4. Each of the filters is configured to selectively pass only a predetermined frequency component.

[0013]

In the hands-free communication device according to the present invention as set forth in claim 1, the first and second microphones convert the transmission voice containing the voice by the acoustic coupling from the speaker to each microphone into an electric signal, and The electric signal is amplified by the first and second transmission signal amplifying means provided corresponding to each microphone. In the direct coupling suppression processing unit, the filter outputs the output signal of the first transmission signal amplifying means and the transfer function representing the ratio of the transfer function of the acoustic direct propagation system from the speaker identified in advance to each microphone directly. After the convolution calculation, the difference between the output signal of the filter and the output signal of the second transmission signal amplification means is obtained by the calculation processing means, and the direct coupling component included in the transmission signal is suppressed.

In the direct coupling suppression processing unit of the hands-free speech communication apparatus according to the present invention as set forth in claim 2, the adaptive filter identifies the characteristic of the output ratio of the microphone, and the first transmission signal amplifying means of the first transmission signal amplifying means. After the convolution operation of the output signal and the identified transfer function, the difference between the output signal of the adaptive filter and the output signal of the second transmission signal amplifying means is obtained by the arithmetic processing means. Then, the oscillation determination means determines the possibility of oscillation based on the output signal of the arithmetic processing means and the input signal of the reception signal amplifier, and the voice switch based on the determination result output by the oscillation determination means. Then, the amount of attenuation is changed to suppress the acoustic coupling component.

According to the third aspect of the present invention, the hands-free communication device according to the present invention corrects the distortion generated in the transmission signal in the output of the direct coupling processing section of the hands-free communication device according to the first or second aspect. It is configured by providing distortion correction means.
In the hands-free communication device according to the present invention as set forth in claim 4, three or more microphones convert the transmission voice including the voice by the acoustic coupling from the speaker to the microphone into an electric signal, and the electric signal is transmitted to each microphone. It is amplified by three or more transmission signal amplifying means provided correspondingly. After the arithmetic processing means obtains the sum of these amplified transmission signals, the voice switch further suppresses the acoustic coupling component contained in the signal obtained by the sum with the application filter.

In the voice switch of the hands-free communication device according to the present invention as set forth in claim 5, the level discriminating circuit provides a coefficient for multiplication to each multiplication processing means based on the signal levels of the reception signal and the transmission signal. Based on these outputs, the first multiplication processing means multiplies the received signal by the coefficient, and the second multiplication processing means multiplies the received signal by the coefficient and outputs the result.

According to the sixth aspect of the present invention, the hands-free communication device according to the present invention selects only the frequency component of the predetermined band of the transmission signal output by the direct coupling suppression processing unit by the filter and outputs it to the voice switch. In the hands-free communication device according to the present invention as set forth in claim 7, a frequency component of a predetermined band of the transmission signal output by each transmission signal amplification means is provided by a filter provided corresponding to each transmission signal amplification means. Only select and input directly to the coupling suppression processing unit.

Here, the acoustic coupling component includes both the direct coupling component and the indirect coupling component included in the transmission signal, and the suppression of the direct coupling component or the acoustic coupling component means the direct coupling component or the acoustic coupling component. Not only when it is completely removed, but also when the direct coupling component or the acoustic coupling component is attenuated.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an example of the construction of a hands-free communication device according to the first aspect of the present invention. This hands-free communication device includes a speaker 30, a first microphone 31a, and a second microphone 3 attached to a housing 3.
1b, an amplification unit 20 that amplifies a reception signal and outputs it to the speaker 30, a first transmission signal amplification unit 21a and a second transmission signal amplification unit that amplifies transmission signals output from the microphones 31a and 31b, respectively. It is composed of the means 21b, the direct coupling suppression processing section 50, and the voice switch 1.

The voice switch 1 is the voice switch circuit described above, which is composed of an analog circuit, like the conventional speech communication apparatus, and suppresses the acoustic coupling component in the transmission signal. The microphones 31a and 31b are arranged so that the distances from the center of the speaker 30 are different, and the respective output signals are amplified by the transmission signal amplifying means 21a and 21b and input to the direct coupling suppression processing unit 50.

The direct coupling processing section 50 has A / D converters 500 and 501 for converting a transmission signal into a digital signal, a filter 502, an arithmetic processing means 503, and an arithmetic processing means 5.
And the D / A converter 504 for converting the output signal of No. 03 into an analog signal. The filter 502 performs a convolution operation of the output signal of the first transmission signal amplifying means 21a after digital conversion and a predetermined transfer function, and the arithmetic processing means 503 performs the convolution operation of the output signal of the filter 502 and the digital conversion result. The difference from the output signal of the second transmission signal amplifying means 21b is obtained.

Since the coefficient of the filter 502 is preset to a value that minimizes the direct coupling component included in the output signal of the arithmetic processing means 503, the direct coupling processing unit 50 includes the transmission signal. It is possible to suppress the generated direct coupling component. Since the direct coupling suppression processing unit 50 is a digital processing device, a digital signal processing device, a so-called DS
It can be realized by P.

An example of the structure for identifying the coefficient of the filter 502 is shown in FIG. The handset terminal of the hands-free communication device is not connected to the line system, the white noise oscillator 6 is connected to the reception signal amplifying means 20 as a signal source, and the arithmetic processing means 50 is connected.
The three output signals e are connected to a filter 502 configured as an adaptive filter. In addition, the microphone 31a,
The coefficient of the filter 502 is identified by using an algorithm that minimizes the signal e after a state in which there is no reflective object around so that 31b does not pick up the indirect coupling component.

As the algorithm, for example, LMS
The case where an algorithm based on the least square method called an algorithm is adopted is shown in the following equation. W (k + 1) = W (k) +2 μeX (k) (1) where W (k) is the coefficient vector of the filter at the sample time k, X (k) is the input vector of the filter at the sample time k, and μ is the step 0-1 representing the size
, And e is the error signal.

Here, the coefficient of the filter 502 is determined by
This is equivalent to identifying the difference in impulse response of the acoustic direct coupling system from the speaker 30 to the microphones 31a and 31b. Therefore, the filter 502 is the speaker 3
It is not necessary to identify the impulse response of the electroacoustic transducer system such as 0, and it is sufficient to identify only the characteristic of the acoustic transmission system until the sound wave output from the speaker 30 reaches the microphones 31a and 31b. The required order is significantly reduced compared to the adaptive filters used in conventional echo cancellers.

This state is shown in FIGS. 3 (a) to 3 (c). 4 (a) and 4 (b) are block diagrams showing an outline of the hands-free communication device for explanation of FIG. 3, and FIG.
4B shows the case where the conventional echo canceller system is adopted, and FIG. 4B shows the case where the system according to the present invention described in claim 1 is adopted. Figure 4 (a)
Only one microphone 31 is provided, the distance from the speaker 30 to the microphone 31 is L1, and the filter Wm1 suppresses the acoustic coupling component of the transmission signal based on the reception signal.

On the other hand, FIG. 4B is provided with two microphones 31a and 31b, the distances from the speaker 30 to the microphones 31a and 31b are L1 and L2, respectively, and the filter Wm2 is a transmission signal from the microphone 31a. The acoustic coupling component of the transmission signal is suppressed based on
When the direct coupling component is suppressed by the conventional echo canceller, the filter Wm1 needs to identify the impulse response h1 itself of the acoustic direct coupling system. The impulse response h1 usually has the characteristics shown in FIG. 3A. For example, when the distance between the speaker 30 and the microphone 31 is about 10 cm, the impulse response length Tm1 is 20 m. It will be about a second.

Therefore, if the sampling period is 8 kHz, the order of the filter Wm1 necessary for identifying the impulse response h1 is 160 or more. On the other hand, in the case of the present invention in which the two microphones 31a and 31b directly suppress the coupled component, the filter Wm2 thereof is as shown in FIG.
As shown in (a) and (b), each microphone 31a,
It is used to identify the difference between the two impulse responses h1, h2 corresponding to 31b. Therefore, the speaker 3
When 0 can be regarded as a point sound source, the optimum impulse response of the filter Wm2 is a delayed impulse as shown in FIG. 3 (c).

Therefore, by arranging the speaker 30 and the microphones 31a and 31b at appropriate positions,
The order required for the filter is significantly reduced compared to the adaptive filter used in the conventional echo canceller, and the performance of the processing speed required for signal processing hardware for howling suppression is It is significantly reduced compared to the echo canceller method.

Further, in the case of an ordinary hands-free speech communication apparatus, the ratio of the direct coupling component and the indirect coupling component contained in the transmission signal is about 10: 1, and the direct coupling component is dominant, and direct coupling suppression is performed. If the direct coupling component can be sufficiently suppressed by the part, howling can be effectively suppressed, and howling can be prevented even if the indirect coupling component fluctuates due to changes in the surrounding environment. It is possible to make a stable call.

Therefore, by sufficiently suppressing the direct coupling component by the direct coupling suppression processing unit, a hands-free speech communication device capable of suppressing the acoustic coupling component in a stable and effective manner is constituted by relatively inexpensive hardware. can do. FIG. 5 shows an example of the configuration of the hands-free communication device according to the second aspect of the present invention. In this hands-free communication device, the direct coupling processing unit 51 has an input signal to the direct coupling suppression processing unit 50 shown in FIG. 1, an adaptive filter 510, a second arithmetic processing unit 511, and a reception signal amplification unit 20. Is provided with an A / D converter 512 for converting the signal into a digital signal, an oscillation determining unit 513, and a D / A converter 514 for converting the output signal of the oscillation determining unit 513 into an analog signal.

The adaptive filter 510 identifies the characteristic of the ratio of the output of the microphone, and performs a convolution operation on the output signal of the first transmission signal amplifying means 21a converted into a digital signal and the identified transfer function. Then, the output signal of the adaptive filter 510 and the second transmission signal amplifying means 21b.
The difference between the output signal and the output signal is obtained by the second arithmetic processing means 511.

The oscillation determining means 513 determines the possibility of oscillation based on the amplitude ratio between the output signal of the arithmetic processing means 511 and the input signal of the reception signal amplifier 20 converted into the digital signal, and the D The / A converter 514 converts the determination signal output from the oscillation determination means 513 into an analog signal and outputs the analog signal to the voice switch 1. The coefficient of the adaptive filter 510 is an adaptive filter that is adaptively updated by an algorithm that minimizes the output signal of the second arithmetic processing means 511, and its order is much lower than that of the filter 502. To do.

Here, when the system is not in the oscillating state and the call is normally made, the two microphones 31 are connected.
There are no distinct peaks in the spectra of the output signals of a and 31b. In this case, in order to reduce the output signal of the arithmetic processing means 511, the adaptive filter 510 needs to uniformly reduce the identification error over a wide band, but is adaptive to the impulse response length of the system to be identified. Filter 5
Since the order of 10 is much lower, the identification error is large and the ratio between the received signal and the output signal of the second arithmetic processing means 511 is large.

On the other hand, in the state immediately before the system oscillates, the spectrum of the received signal and the output signals of the microphones 31a and 31b have sharp peaks in a narrow band.
In this case, the amplitude of the time waveform of the error signal can be significantly reduced by reducing the identification error in a narrow band near the frequency where the peak exists. It is possible to identify the amplitude and phase characteristics in a very limited band of the frequency response of the system to be identified by the adaptive filter 510 having a low order, and the amplitude of the peak in the spectrum of the output signal of the arithmetic processing means 511 can be identified. The ratio between the received signal and the output signal of the second arithmetic processing means 511 becomes smaller.

Therefore, the oscillation judging means 513 monitors the amplitudes of the received signal and the output signal of the arithmetic processing means 511, and when the ratio is larger than a predetermined value, it is judged that it is in a normal state. However, if the ratio is smaller than a predetermined value, it can be determined that the state is just before oscillation. The voice switch 1 is based on the determination result from the oscillation determination means converted into an analog signal,
If the system is in the state just before oscillating, change the insertion loss amount to a value larger than that in the normal state.

That is, the oscillation determining means can detect the state immediately before oscillation, and when it is determined to be in the oscillation state, the insertion loss amount in the voice switch is changed to a large value to cause oscillation. Can be prevented. In particular, even when the acoustic indirect coupling system suddenly changes and the indirect coupling component increases, oscillation can be prevented and stable operation can be performed.

FIG. 6 shows an example of the configuration of the hands-free communication device according to the third aspect of the present invention. In this hands-free communication device, the distortion correction filter 8 for correcting the output of the arithmetic processing unit 503 of the direct coupling suppression processing unit 50 for the utterance voice included in the transmission signal, that is, the component of the voice of the caller. Is provided, and the output signal of the distortion correction filter 8 is input to the audio switch 1 via the D / A converter 504.

From the caller, two microphones 31a,
Since the distance to 31b is almost equal, the uttered voice, that is,
Two microphones 31 depending on the voice of the caller
The sound pressures generated at the positions a and 31b have the same phase and the same amplitude. This sound pressure is Ps, and two microphones 3
1a and 31b and two signal amplifier means 21a and 2
If the characteristics of 1b are equal to each other, and the transfer function of the microphone, the transmission signal amplifying means, and the A / D converter is H and the transfer function of the filter is Wd, the utterance included in the output signal of the direct coupling suppression processing unit 51. The voice component is H
(1-Wd) Ps.

In the frequency characteristic of the transfer function H, the amplitude is substantially constant in the speech band and the phase is a linear delay characteristic, so that H can be regarded as a simple delay element. However, the frequency characteristic of the transfer function Wd depends on the shape and size of the speaker, the shape and size of the casing of the communication device, and
Since it is determined by the arrangement of the speaker and the microphone, it is generally complicated in the call band. That is, (1-Wd) is not just a delay characteristic.

Therefore, the direct coupling suppression processing section 50 'causes distortion in the uttered voice component. The distortion correction filter 8 is a filter for correcting this distortion,
In this embodiment, it is configured as a digital filter, but it may be simply configured by an analog element. This distortion correction filter corrects the distortion generated in the speech component by the direct coupling suppression process,
The call quality can be improved.

FIG. 7 shows an example of the configuration of the hands-free communication device according to the fourth aspect of the present invention. This hands-free communication device includes a speaker 30 attached to the housing 3 and three or more microphones 31a to 31n, an amplification means 20 for amplifying a received signal and outputting the amplified signal to the speaker 30.
It is composed of transmission signal amplification means 21a to 21n for amplifying the transmission signals output from the microphones 31a to 31n, a direct coupling suppression processing section 52 composed of the arithmetic processing means 520, and the voice switch 1.

Each of the above microphones 31a to 31n
Are arranged so that the distances from the center of the speaker 30 are different, and the respective output signals are transmitted signal amplifying means 21a.
21n are amplified, added by the arithmetic processing means 520, and input to the voice switch 1. Since the distance from the speaker 30 to each of the microphones 31a to 31n is different, the phase of the direct coupling component of the sound wave picked up by each of the microphones 31a to 31n differs depending on each microphone. On the other hand, the microphones 31a-3 from the caller
The distance from 1n is from the speaker 30 to the microphone 3
Since the distance is sufficiently longer than the distance from 1a to 31n, the phases of the voices emitted by the callers included in the transmission signals output from the microphones 31a to 31n are equal.

Therefore, each of the transmission signal amplifying means 21a-2
When the output signals of 1n are added by the arithmetic processing means 520, only the directly coupled components are canceled and attenuated, while the voice component of the caller is amplified. In this way, since the direct coupling component can be suppressed by the direct coupling suppression processing unit 52, the acoustic coupling processing unit 1 mainly suppresses the indirect coupling component.

The above microphones 31a to 3a are used.
By arranging 1n in a row in the radial direction of the vibrating surface of the speaker 30, it is possible to give the sound collecting characteristic directivity. FIG. 8 shows an example of the configuration of the hands-free communication device according to the sixth aspect of the present invention. In this hands-free communication device, a filter 7 is provided at the output of the arithmetic processing means 503 of the direct coupling processing unit 50 shown in FIG. 1, and the output signal of the filter 7 is a voice switch via a D / A converter 504. Input to 1.

The filter 7 is a digital filter that selectively passes only a predetermined frequency component, which is predetermined as a band to be transmitted, of the transmission signal output from the arithmetic processing means 503, and the predetermined frequency. Is determined in consideration of the human voice band, the characteristics of the speaker, the call band, and the like. Since the filter 7 removes components other than the necessary frequency components, it is possible to prevent the voice switch 1 from malfunctioning due to the influence of noise other than the frequency band picked up by the microphone, and improve the call quality. be able to.

FIG. 9 shows an example of the structure of the hands-free communication device according to the present invention described in claims 5 and 7. This hands-free communication device is provided with filters 7a and 7b at the outputs of the transmission signal amplifying means 21a and 21b of the hands-free communication device shown in FIG. 1, and the output signals of the filters 7a and 7b are directly coupled and suppressed by the suppression unit 52. Is configured to be the input of.

The voice switch 1'includes a first multiplication processing means 11 for multiplying a transmission signal and a predetermined coefficient, and a second multiplication processing means 12 for multiplying a reception signal and a predetermined coefficient.
And a level determination section 10 for outputting the above-mentioned respective coefficients to be multiplied in the multiplication processing means 11, 12 based on the signal levels of the transmission signal and the reception signal, and converting the transmission signal into an analog signal. The D / A converter 13, the A / D converter 14 for converting the received signal into a digital signal, and the D / A converter 15 for converting the received signal into an analog signal constitute a digital signal processing means.

Like the filter 7, the filters 7a and 7b are filters that selectively pass only a predetermined frequency component of the transmission signal, which is predetermined as a band to be transmitted, and are analog filters. Composed.
Therefore, similarly to the filter 7, due to the influence of noise or the like other than the frequency band picked up by the microphone,
It is possible to prevent the voice switch 1 from malfunctioning and improve the call quality.

Also, as a preprocessing of A / D conversion,
A low pass filter is required to prevent aliasing, but the output signals of the filters 7a and 7b are A /
If the sampling frequency fs of the D converters 500 and 501 is not included, that is, if the frequency component of fs / 2 or more is not included,
The low-pass filter is unnecessary. This state is shown in FIG.
It shows in (a)-(e). 10A is an example of the frequency spectrum of the output signals of the microphones 31a and 31b, FIG. 10B shows the frequency characteristics of the low-pass filter, and FIG. 10D shows the frequency of the filters 7a and 7b. It is a figure which shows an example of a characteristic.

The output signal of the low-pass filter in this case is shown in FIG. 10 (c), and the output signals of the filters 7a and 7b are shown in FIG. 10 (e). As shown in the frequency spectrum of FIG. 10E, the output signals of the filters 7a and 7b are
If only the components of fs / 2 or less are provided, the filters 7a and 7b also have a function of preventing aliasing, and a low-pass filter becomes unnecessary.

Next, in the voice switch 1 ', the level discriminating unit 10 monitors the signal levels of the reception signal and the transmission signal, and if the reception signal is higher than the transmission signal, the multiplication of the transmission signal is performed. Decrease the coefficient, or increase the multiplication coefficient of the reception signal, decrease the multiplication coefficient of the reception signal when the level of the transmission signal is higher than that of the reception signal, or increase the multiplication coefficient of the transmission signal. increase. That is, the same operation as an analog voice switch can be performed by digital signal processing.

Therefore, the direct coupling suppression processing unit 52 and the voice switch 1'are both a digital signal processing device, that is, D.
This can be realized by SP, and the number of parts can be reduced as compared with the case of using an analog voice switch, and a hands-free communication device capable of suppressing howling can be provided at low cost. An experiment for confirming the effect of the present invention and its result will be described below. FIG. 11 shows the direct coupling suppression characteristics when the conventional echo canceller method is used, and FIG. 12 shows the direct coupling suppression characteristics when the method according to the present invention is used. The configuration of the apparatus used for the measurement of FIG. 11 is shown in FIG. 13, and the configuration of the apparatus used for the measurement of FIG. 12 is shown in FIG.

The conventional apparatus shown in FIG. 13 is an echo canceller composed of an adaptive filter, a subtracting means, an A / D converter and a D / A converter, and an acoustic coupling component contained in the output signal of the subtracting means. The filter coefficient is adaptively updated so that is minimized. On the other hand, the device according to the present invention shown in FIG. 14 has the same configuration as that of the device shown in FIG. 1 and, as already explained, the filter of the filter so that the direct coupling component contained in the output signal of the subtracting means is minimized. Coefficients have been adaptively updated.

The output signals of the white noise oscillator are supplied to the A / D converter 13 as the reception signals in the acoustically updated room in which the two devices thus adaptively updated. At that time, the D / A converter 1
The transmission signal output from No. 4 was observed by an FFT analyzer. 11 and 12 configure the above-described adaptive filter of each device as a filter having a tap number of 16 taps, 32 taps, and 64 taps, respectively, and adaptively update the coefficient to converge the coefficient and then output the output. The frequency characteristic of the direct coupling component included in the speech signal is shown. The sampling frequency was 8 kHz, and the cut-off frequency of the aliasing prevention low-pass filter was 3 kHz.

As can be understood from these measurement results,
It has been shown that the filter order required to obtain the same degree of direct coupling suppression effect is reduced to ¼ or less of the conventional method by the method according to the present invention.

[0057]

In the hands-free communication device according to the first aspect of the present invention, the filter of the acoustic coupling suppression processing unit is
Since it suffices to identify the difference between the impulse responses of the acoustic coupling system from the speaker to the two microphones, the order of the filter is significantly reduced compared to the conventional echo canceller type communication device that identifies the impulse response of the entire acoustic coupling system. can do. That is, the performance of the calculation processing speed required for signal processing hardware for howling suppression is
It is significantly reduced compared to the conventional echo canceller system.

Further, since the direct coupling component occupying most of the acoustic coupling component can be sufficiently removed by the direct coupling suppression processing section, even when the indirect coupling component fluctuates due to changes in the surrounding environment and the like. , Stable processing can be performed. Therefore, it is possible to inexpensively provide a hands-free communication device that enables simultaneous two-way communication without feeling disconnection or blocking reception during a call.

The hands-free communication device according to the second aspect of the present invention has a function of determining the possibility of oscillation,
When the system is in a state immediately before oscillating, oscillation can be prevented by changing the insertion loss amount in the voice switch to a large value. In particular, even when the acoustic indirect coupling system is suddenly changed and the indirect coupling component is increased, oscillation can be prevented and stable operation can be performed, which is stable and effective. It is possible to provide a hands-free communication device capable of suppressing howling.

In the hands-free communication device according to the third aspect of the present invention, by providing the distortion correction means, it is possible to correct the distortion of the uttered voice caused by the direct coupling suppression process, and improve the communication quality. You can In the hands-free communication device according to the fourth aspect of the present invention, since the suppression processing can be performed by adding the output signals of three or more microphones having different distances from the speaker, howling is suppressed by a simple configuration. be able to. That is, it is possible to inexpensively provide a hands-free communication device capable of suppressing howling.

In the hands-free communication device according to the fifth aspect of the present invention, since the voice switch is realized by digital signal processing, both the direct coupling suppression processing unit and the voice switch are digital signal processing devices, so-called DS.
It can be realized by P, and the number of parts can be reduced as compared with the case where an analog voice switch is used, and a hands-free communication device capable of suppressing howling can be provided at low cost.

In the hands-free speech communication apparatus according to the sixth aspect of the present invention, the band of the acoustic coupling component to be suppressed by the voice switch can be limited by providing the analog filter that passes only the predetermined band. For,
It is possible to reduce the feeling of disconnection during a call and the occurrence of call blocking. Further, ambient noise other than the call band can be cut, the S / N ratio can be improved, and the call quality can be improved.

A hands-free communication device according to a seventh aspect of the present invention has the same effects as described above, and
Since it is provided in the preceding stage of the A / D converter, it can function as an anti-aliasing filter for preventing aliasing, and the low-pass filter normally used as an anti-aliasing filter can be reduced, which is more effective. It is possible to inexpensively provide a hands-free communication device that effectively suppresses acoustic coupling components and improves the communication quality.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration example of a hands-free communication device according to the present invention as set forth in claim 1;

FIG. 2 is a diagram illustrating a configuration example for identifying a coefficient of a filter of a direct coupling suppression processing unit.

FIG. 3 is a diagram showing an example of an impulse response of a filter.

FIG. 4 is a block diagram showing an outline of a hands-free communication device for explaining FIG.

FIG. 5 is a diagram showing a configuration example of a hands-free communication device according to the present invention as defined in claim 2;

FIG. 6 is a diagram showing a configuration example of a hands-free communication device according to the present invention as defined in claim 3;

FIG. 7 is a diagram showing a configuration example of a hands-free communication device according to the present invention as defined in claim 4;

FIG. 8 is a diagram showing a configuration example of a hands-free communication device according to the present invention according to claim 6;

FIG. 9 is a diagram showing a configuration example of a hands-free communication device according to the present invention described in claims 5 and 7.

FIG. 10 is a diagram showing an example of a frequency spectrum of a transmission signal and a frequency characteristic of a filter.

FIG. 11 is a diagram showing an observation result when a conventional method is used.

FIG. 12 is a diagram showing an observation result when the method according to the present invention is used.

13 is a diagram showing a configuration of an apparatus used for the observation of FIG.

14 is a diagram showing a configuration of an apparatus used for observation in FIG.

FIG. 15 is a block diagram showing a configuration example of a voice switch.

FIG. 16 is a diagram showing characteristics of a voice switch.

[Explanation of symbols]

 1, 1 '... Voice switch 50, 50' 51, 52 ... Direct coupling suppression processing unit 502, 7, 7a, 7b ... Filter 510 ... Adaptive filter 3 ... Case 30 .. -Speaker 31a, 31b ... Microphone 20 ... Received signal amplification means 21a, 21b ... Transmitted signal amplification means 503, 511 ... Calculation processing means 513 ... Oscillation determination means 8 ... Distortion correction means

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kako Takura 1048, Kadoma, Kadoma-shi, Osaka Matsushita Electric Works, Ltd.

Claims (7)

[Claims] 1. A howling caused by acoustic coupling from a speaker to a microphone in which a speaker for converting a received signal into a voice and outputting the same and a microphone for converting a transmitted voice into an electric signal and outputting the same are mounted in the same casing. In a communication device equipped with a voice switch for suppressing, the microphone is composed of a first microphone arranged in a position close to the speaker and a second microphone arranged in a position distant from the speaker. First transmitting signal amplifying means for amplifying the output signal, second transmitting signal amplifying means for amplifying the output signal of the second microphone, and receiving signal amplifying means for amplifying the input signal to the speaker are provided. A direct connection that suppresses the voice component that is included in the transmitted signal and that goes directly to the microphone through the speaker. A suppression processing unit is provided in the preceding stage of the voice switch, and the direct coupling suppression processing unit is a sound direct propagation system from the speaker previously identified with the output signal of the first transmission signal amplifying means to each microphone directly. And a transfer function representing a ratio of transfer functions of the filter, and a processing unit for calculating a difference between the output signal of the filter and the output signal of the second transmission signal amplifying unit. Characteristic hands-free communication device. 2. A characteristic of the ratio of the output of the microphone is identified in the direct coupling suppression processing section, and a convolution operation is performed on the output signal of the first transmission signal amplifying means and the identified transfer function. Based on an adaptive filter, an arithmetic processing means for obtaining a difference between an output signal of the adaptive filter and an output signal of the second transmission signal amplifying means, and an output signal of the arithmetic processing means and an input signal of the reception signal amplifier. And an oscillation determination means for determining the possibility of oscillation is provided, and the voice switch is configured as a voice switch that changes its attenuation amount based on the determination result output by the oscillation determination means. The hands-free communication device according to claim 1. 3. The hands-free communication device according to claim 1, wherein the output of the direct coupling suppression processing unit is provided with distortion correction means for correcting distortion generated in the transmission signal. 4. A howling caused by acoustic coupling from a speaker to a microphone in which a speaker for converting a received signal into a voice and outputting the same and a microphone for converting a transmitted voice into an electric signal and outputting the same are mounted in the same casing. In a communication device having a voice switch for suppressing the above, the microphone is composed of three or more microphones arranged at different distances from a speaker, and is provided corresponding to each of the above microphones. It is equipped with three or more transmitting signal amplifying means for amplifying the output signals respectively and receiving signal amplifying means for amplifying the input signal to the speaker, which is included in the transmitting signal and directly to the microphone via the speaker. A direct coupling suppression processing unit that suppresses the voice component that wraps around the Direct coupling suppressing unit, a hands-free device, characterized in that it is constituted by a calculation processing means for obtaining the sum of the output signal of the transmission signal amplifying means. 5. A first multiplication processing means for the voice switch of the acoustic coupling suppression processing section to multiply a received signal and a predetermined coefficient, and a second multiplication to multiply a transmission signal and a predetermined coefficient. Processing means, based on the signal level of the reception signal and the transmission signal,
The hands-free communication device according to any one of claims 1 to 4, which is configured by a digital signal processing unit including a level determination unit that outputs each coefficient multiplied by the multiplication processing unit. 6. A filter for selectively passing only a predetermined frequency component, which is predetermined as a band to be transmitted, of the output transmission signal is provided at the output of the direct coupling suppression processing unit. The hands-free communication device according to any one of claims 1 to 5. 7. The output of the transmission signal amplifying means is predetermined as a band to be transmitted, of the output transmission signal,
The hands-free communication device according to any one of claims 1 to 5, further comprising a filter that selectively passes only a predetermined frequency component.