JPH0630089A - Telephone set - Google Patents

Abstract

PURPOSE:To obtain a telephone set capable of having a smooth conversation between a talker and an opponent by eliminating an ambient tone component from the reception signal of a transmitter. CONSTITUTION:A microphone 15 and the transmitter 8 are arranged at different positions, and they receive ambient tones on different routes, and a DSP(digital signal processor) 20 obtains and sets a transfer function K between an ambient sound source and the transmitter 8 former of which minimizes difference of two signals within prescribed period of time according to the start of the operation of transmission and reception. When the talker emits a voice, a transmission signal from which the ambient tone component included in the reception signal from the transmitter 8 is eliminated by a spurious ambient signal based on the transfer function K at that time is attenuated by an attenuator 7 and it is outputted to a speaker 6 as a sidetone, and also, a transmission/reception part 3 transmits it to the opposite side.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a telephone, and more particularly to a telephone for reducing ambient sound and transmitting a voice of a talker.

[0002]

2. Description of the Related Art FIG. 4 is a schematic diagram of a conventional portable telephone. In the figure, 1 is a numeric keypad for setting the telephone number of the other party, 2 is a hook button that is pressed when calling the other party, or when there is a call from the other party, 3 is at least a telephone call from the ten key 1 It is a transmitting / receiving unit that transmits the number to the other party, recognizes the voice of the talker and transmits it as a transmission signal, and receives the transmission signal from the other party.

Reference numeral 5 is an amplifier for amplifying the transmission signal of the other party from the transmitting / receiving section 3, 6 is a speaker for converting the transmission signal of the other party from the amplifier 5 into voice, and 8 is the voice of the sender. Then, a transmitter (hereinafter, simply referred to as a microphone) for converting into a sound reception signal, an amplifier 9 for amplifying the sound reception signal from the microphone 8 and outputting the amplified sound reception signal to the transmitter / receiver 3 is connected in parallel to the microphone 8 and the speaker 6. It is an attenuator that is connected and attenuates the sound reception signal from the microphone 8 and outputs it as side sound to the speaker 6 side.

The operation of the portable telephone configured as described above will be described below. Generally, when a sound is emitted in the air, the sound N is transmitted based on the transfer function K of the space.
Therefore, for example, when a speaker makes a phone call in a place where there is noise, when the sound N of the ambient sound source per unit time is received, the sound N and the transfer function K are superimposed on the sound receiving side (NK It is known that the sound is received as a sound indicated by the function.

On the basis of such a sound propagation property, for example, it is assumed that the talker brings his mouth close to the transmitter 2 and the transfer function K between the talker and the microphone 8 can be ignored. .
For example, when the talker makes a call in a place where there is noise or the like, the ambient sound (NK) and the talker's voice S are received by the microphone 8, converted into an electric signal and output as a sound reception signal.

The received sound signal becomes the voice S of the sender and the ambient sound (NK), is amplified by the amplifier 4, and is output to the transmitter / receiver 3. The transmitter / receiver 3 receives this sound reception signal (S + N
K) is transmitted as a transmission signal to the other party, and the transmission signal of the other party is output to the speaker 6 via the amplifier 5. Further, the attenuator 10 receives the sound receiving signal (S +
NK) was attenuated and output to the speaker 1 side as a side sound.

[0007]

In a conventional telephone, an attenuator is connected between a speaker and a transmitter to output the speaker's own voice to the speaker side at a low level, and to output the speaker as a side tone. Notifying the user that his / her voice is being received reliably, the sound receiving signal from the transmitter is transmitted to the other party, and the other party's transmitting signal is received and output to the speaker. However, when the ambient sound is large, the ambient sound is superimposed on the voice of the speaker and is received by the microphone, and is output to the speaker via the attenuator, so that the S / N is deteriorated and most of the side sound from the speaker is generated. It becomes the ambient sound. Therefore, there is a problem that the speaker determines that the noise due to the ambient sound is large and that the speaker himself needs to increase the voicing level to speak.

Further, since the transmitter converts the sound reaching the sound receiving surface into an electric signal and outputs the electric signal, the speaker's voice becomes a transmitter signal on which the ambient sound is superimposed, and is transmitted to the other party. Sent.
Therefore, when received by the other party, the other party hears the voice of the talker together with the ambient sound of the talker, and if the ambient sound of the talker is very loud, There was a problem that it became difficult to hear the voice.

The present invention has been made in order to solve the above problems, and removes the ambient sound component from the sound reception signal of the transmitter to enable a telephone for both the sender and the other party to have a smooth conversation. Aim to get.

[0010]

A telephone according to the present invention is provided with a microphone which receives an ambient sound and outputs it as an ambient sound signal, and is arranged at a position different from that of the microphone. A transmitter that receives sound and outputs it as a sound reception signal, and with the start of the transmission or reception operation,
Within a predetermined time, a pseudo ambient sound signal is generated by the ambient sound signal from the microphone and a predetermined transfer function, and the pseudo ambient sound signal is compared with the sound reception signal by the ambient sound from the transmitter, and both The transfer function is updated and set so that the difference between the signals is minimized, and after a lapse of a predetermined time, the pseudo ambient sound signal component at that time is removed from the sound reception signal output from the transmitter, and the An arithmetic unit that outputs a transmission signal, an attenuator that outputs the transmission signal from the calculation unit to a speaker, and a transmission signal from the calculation unit that is transmitted to the other party and receives the transmission signal from the other party. The transmitter / receiver outputs to a speaker.

[0011]

In the present invention, the microphone receives the ambient sound and outputs it as an ambient sound signal. Since the transmitter is arranged at a position different from that of the microphone, before the speaker speaks, the ambient sound received through a route different from that of the microphone is output as a sound reception signal. When the speaker utters a voice, the ambient sound and the voice of the speaker are superimposed and the received sound signal is output.

Then, the arithmetic means generates a pseudo ambient sound signal by the ambient sound signal from the microphone and a predetermined transfer function within a predetermined time with the start of the transmitting or receiving operation, and the pseudo ambient sound signal. And the received signal due to the ambient sound from the transmitter. Then, the transfer function is updated so that the difference between the two signals is minimized, a pseudo ambient sound signal is generated by the ambient sound signal and the updated transfer function, and the transmitter is generated each time the pseudo ambient signal is generated. The received signal due to the ambient sound from is compared and the transfer function that minimizes the difference between the two signals is set.

Next, after a lapse of a predetermined time, a pseudo ambient sound signal is generated by the ambient sound signal from the microphone and the set transfer function, and from the sound reception signal output from the transmitter, the pseudo ambient sound signal at that time is generated. The component of the sound signal is removed, and it is output as a transmission signal. Then, the attenuator attenuates the transmission signal from which the component of the pseudo ambient sound signal is removed and outputs it as a side sound to the speaker. Further, the transmitting / receiving unit transmits the ambient sound signal from which the component of the pseudo ambient sound signal has been removed, to the other party.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a schematic block diagram showing an embodiment of a telephone according to the present invention. In the figure, 1 to 10 are the same as those in FIG. Reference numeral 15 is a microphone that collects ambient sound and outputs it as an ambient sound signal. The microphone 15 has directivity so as not to collect the voice generated by the speaker.

Reference numeral 16 denotes an amplifier for amplifying the ambient sound signal from the microphone 15, and 17 denotes quantizing the ambient sound signal from the amplifier 16 based on a predetermined sample signal, and outputting ambient sound data of a predetermined number of bits for each sample. An A / D converter 18 is an A / D converter that quantizes the sound reception signal from the amplifier 9 based on a predetermined sample signal and outputs sound reception data of a predetermined number of bits for each sample.

Reference numeral 20 denotes at least the ambient sound data from the A / D converter 17 and the sound reception data from the A / D converter 18, and removes the ambient sound data component contained in the sound reception data to output a transmission sound. DSP (Digital) that outputs as data
tal Signal Processor). The DSP 20 starts a transmitting or receiving operation by pressing the hook button 2, and within a predetermined time, the microphone 1
Pseudo ambient sound data is generated by the ambient sound data from 5 and a preset transfer function from the ambient sound source to the microphone 8, and the pseudo ambient sound data and the sound reception data by the ambient sound from the microphone 8 are generated. By comparison, the transfer function is updated and set so that the difference between the two is minimized, and after the elapse of a predetermined time, the component of the pseudo ambient sound data at that time is removed from the sound reception data output from the microphone 8.

Reference numeral 21 is a D / A converter that converts the transmission sound data from the DSP 20 into analog and outputs it as a transmission signal, 22 is an amplifier that amplifies the transmission signal from the D / A converter 21, and 23 is It is an adder that synthesizes the transmission signal from the attenuator 7 and the transmission signal of the other party output from the transmission / reception unit 10 and outputs the synthesized signal to the speaker 6 side. The attenuator 7 is connected to the output side of the buffer 22 and the adder 23,
The transmission sound data of the DSP 20 is attenuated and output as a side sound signal to the speaker 6 side.

FIG. 2 is a block diagram showing a schematic function of the DSP according to the present invention. However, the amplifier is omitted. In the figure, 2 to 20 are the same as those in FIG. 31a to 31
m is a delay element group. This delay element group is composed of a program, and is accumulated each time ambient sound data for each sample from the A / D converter 17 is input, and for each next ambient sound data for each sample input. The ambient sound data for each is output.

Reference numerals 33a to 33m are linear prediction element groups.
This linear prediction element group is configured by a program, and based on the weighted coefficients, the delay element 31a
-Ambient sound data for each sample from the delay element 31m is converted and output. The addition means 35 is addition means 3
5 is configured by a program, and the linear prediction element 33a
Each time the output signal from the linear prediction element 33m is input, it is added and output as pseudo ambient sound data for each sample. 36 is an adding means. The adding means 36 is composed of a program and outputs a difference between the sound reception data for each sample from the microphone 8 and the pseudo ambient sound data for each sample via the A / D converter 18.

Reference numeral 38 is a transfer function calculating means. The transfer function calculating means 38 is configured by a program, receives the difference signal for each sample from the adding means 36, and calculates the transfer function of the space from the ambient sound source to the microphone 8 based on the vector component of the difference signal. Based on the obtained transfer function, the coefficient of each linear prediction element for each sample is obtained by a well-known adaptive algorithm and set in the corresponding linear prediction element, and it is notified that the difference signal has converged within a predetermined time. And a transfer function calculating means for stopping the calculation of the transfer function.

The adaptive algorithm is to correct the coefficient so as to minimize the root mean square error based on the steepest descent method. For example, the LMS in which the corrected part is normalized by the norm of the vector of the input e (t) is used. (Least Mean
Square), RLS (Recursi) equivalent to a Kalman filter when the state variable to be estimated is time invariant
ve Least Square). 38 is a judgment means. The determination means 38 is the addition means 3
The output difference signal 6 is stored each time it is input, and it is determined whether the output difference signal has converged. If it is determined that the output difference signal has converged, the transfer function calculation means 38 stops calculating the transfer function and , Stop working.

That is, when the hook button 2 is pressed, the DSP 20 transmits between the ambient sound source and the microphone 8 so that the difference between the desired signal of the talker's voice and the pseudo ambient sound at the time t is minimized. An adaptive filter is constructed that obtains and updates a function, and updates each coefficient with each update.

FIG. 3 is an external view of a telephone according to an embodiment of the present invention. In the figure, 1 to 15 are the same as above,
40 is a display unit. As shown in the figure, in particular, the microphone 15 is separated from the microphone 8 and is attached to the upper portion of the speaker 1 so that it can be expanded and contracted so as not to capture the voice of the speaker.

The operation of the telephone configured as described above will be described below. For example, assume that the talker is in a place with heavy traffic. When the talker makes a call using a portable mobile phone as shown in FIG. 3, it is common to press the hook button 2 and then speak into the microphone 8 for 1 second. It takes some time, and the transfer function of space is obtained during this period. The telephone of the present invention starts the operation of the present invention by the hook signal output by pressing the hook button 2.

For example, in response to the input of a hook signal output when the hook button 2 is pressed, each unit starts its operation, and the microphone 15 converts the ambient sound N from the ambient sound source into an electric signal to Output as a sound signal. In this case, assuming that the transfer function F between the ambient sound source and the microphone 15 is F, the ambient sound signal from the microphone 15 is FN. At this time, the microphone 8 receives the ambient sound through a route different from that of the microphone 15 because the microphone is not yet uttered by the speaker and is arranged at a position different from the microphone 15. Therefore, the microphone 8 uses the ambient sound signal K based on the transfer function K of the space from the ambient sound source to the microphone 8.
Output N.

Next, the A / D converter 17 quantizes the ambient sound signal FN based on a predetermined sample signal and outputs it as ambient sound data FN. Further, the A / D converter 18 quantizes the transmission signal from the microphone 8 based on a predetermined sample signal and outputs it as sound reception data. Also, DSP
The transfer function calculating unit 38 of 20 sets the prediction elements a _{1 to} a _n in the linear prediction elements 33 a to 33 _n in response to the input of the hook signal.

Next, the DSP 20 outputs 1 of the A / D converter 17.
The ambient sound data for each sample is input as u (t).
This t is the current time at the time of sampling. Then 1
With the input of the ambient sound data u (t) for each sample, the delay element 31a has the previous digital ambient sound data u (t-1) of one sample, and the delay element 31b has the ambient sound data u (two samples before). t-2), the delay element 31m outputs the digital ambient sound data u (tm) m samples before. However, since the delay elements 31a to 31m are set to 0 immediately after the hook button 2 is turned on, the output signal is 0. That is, the delay elements 31a to 31m are sequentially shifted each time ambient sound data u (t) for each sample is input, so that m samples before the ambient sound data u (t) currently input. The digital ambient sound data of is output.

Next, the linear predictive element 33a transforms based on the weighted predictive coefficient a ₁ set every time one sample of ambient sound data u (t) is input, and the linear predictive element 33b to linear predictive element 33b. element 33m is converted based on the prediction element _a 2 ~a _n. However, immediately after pressing the hook button 2,
Since the outputs from the delay elements 31b to 31m are 0, the outputs from the linear prediction elements 33b to 33m are 0.

Next, the adding means 35 adds the output signals of the prediction elements 33a to 33m. That is, the adding means 35 predicts the current 1-sample ambient sound signal by adding the 1-sample past output signal based on the weighted prediction coefficient. This predicted ambient sound data is pseudo ambient sound data y for one sample.
(T). Therefore, the pseudo ambient sound data y (t) is expressed by the following equation 1.

[0030]

[Equation 1]

M in this equation is the number of taps and is the last number of the delay element. Further, h (m) is the calculated transfer function H. m is the sample number and t is the sample time. Further, the A / D converter 18 samples the sound reception signal from the microphone 8 with a predetermined sample signal and outputs a digital transmission signal for each sample to the DSP 20. D
SP20 sends the transmission data (KN) for each sample
Input as (t). Then, the adding means 36 outputs the difference between the transmission data d (t) at one sample and the ambient sound data y (t) at one sample as the output difference signal e (t). That is, one sample of pseudo ambient sound data (HN = y
(T)) and the transmission data from the microphone 8 (KN = d
Since the output difference signal e (t) from (t)) is output,
It becomes the formula shown in.

[0032]

## EQU00002 ## e (t) = d (t) -y (t) ... (2)

Next, the transfer function calculating means 38 is added by the adding means 3.
Every time the output difference signal e (t) from 6 is input, the microphone 8 is output based on the vector component for each output difference signal e (t).
The transfer function H (t) of the space up to is obtained. Then, every time the transfer function H (t) is obtained, the linear prediction element 3 is calculated by a well-known adaptive algorithm based on the transfer function H (t).
3a to linear prediction element 33m, weighting coefficient a _{1 to} a
m is obtained and set (updated) in the corresponding linear prediction elements 33a to 33m.

The determining means 39 stores the output signal e (t) each time it is output, and determines whether or not the output signal e (t) is converged based on the degree of change from the previous output signal e (t). If it is determined that they have converged, the operation of the transfer function calculation means 38 is stopped. That is, the DSP 20 collects the ambient sound before the speaker produces a voice by pressing the hook button 2, estimates the transfer function K between the ambient sound source and the microphone 8 for each sample, and estimates the transfer function K. The pseudo ambient sound data is generated by the transfer function H, and the pseudo ambient sound data and the microphone 8 are generated.
And the received sound data due to the ambient sound are compared, and the transfer function H is updated and set so that the difference between the two is minimized. Then, for example, when the talker makes a voice, it is received by the microphone 8 as a sound (S + KN) to which the ambient sound (KN) is added, and is output to the A / D converter 18 as a talk signal. The A / D converter 18 samples the transmission data (S + K
N) is output.

Further, the microphone 15 uses the ambient sound signal FN.
And the A / D converter 17 samples in the same manner as above and outputs ambient sound data FN. Then, the delay element 31
a to ambient sound data u (t) shifted by each sample by the delay element 31m, and the linear prediction element 33a obtains the current ambient sound data u (t), and sets the transfer function H (= h _The conversion is performed by the coefficient a ₁ according to _n (n = 1 to m), and the linear prediction element 33b to the linear prediction element 33m convert the previous ambient sound data u (t−) for each sample.
Input m) and convert. Then, each time it is added by the adding means 35, it is output as pseudo ambient sound data y (t) corresponding to the ambient sound currently input. In this case y
(T) is HFN.

Next, the adding means 36 adds S + KN which is the sound reception data d (t) of the microphone 8 and the pseudo ambient sound data y (t).
Is output as output difference data e (t). Therefore, if y (t) = HFN = Y, then the DSP
The output Z of 20 is given by:

[0037]

## EQU00003 ## Z = (S + KN) -Y = S + (K-H) N ... (3) where Y = HN

According to these three equations, when approximately F = H, the output signal Z of the DSP 20 is the desired signal S = S. In this case, since the transfer function H is obtained in advance so that the ambient sound received by the microphone 8 and the ambient sound received by the microphone 2 coincide with each other, K is almost equal to H, so that the voice signal S of the speaker is obtained.
Is the desired signal.

Next, the sound reception data Z (= S) of the transmitter from the DSP 20 is output as the transmission data, and the D / A 21 is output.
Is converted to analog and output via the amplifier 22. Then, the attenuator 10 attenuates the transmission data Z (= S) and outputs it as sidetone to the speaker 6 side. Therefore, the talker can hear only his own voice as the sidetone, and the other party can also clearly hear the voice of the talker. In the above embodiment, the microphone is provided above the speaker, but in the case of a mobile phone, it may be provided on the back surface of the main body. Further, when the output difference signal converges, the calculation of the transfer function H is stopped.
The time of s may be counted and the calculation of the transfer function H may be stopped when 100 ms has elapsed.

Further, in the above embodiment, A / D or D / A
Is attached before and after the DSP, an arithmetic means having the functions of A / D and D / A may be used. Furthermore, in the above-mentioned embodiment, a mobile telephone for mobile use has been described, but a general telephone equipped with a handset may be used.

[0041]

As described above, according to the present invention, the microphone and the transmitter are arranged at different positions, and ambient sounds from ambient sound sources that arrive on different routes are received to transmit or With the start of sound receiving operation, within a predetermined time, the transfer function between the transmitters was set from the ambient sound source such that the difference between the two signals was the smallest, and was set, and the speaker uttered a voice. In this case, the pseudo ambient sound signal based on the transfer function at that time attenuates the ambient sound component included in the received sound signal from the transmitter, attenuates it, and outputs it as a side sound to the speaker. Since it is transmitted to the side, the effect that the speaker can hear only the speaker's own sound as the sidetone and the other party can clearly hear the speaker's voice is obtained. ing. Therefore, even if the ambient sound is loud, the intelligibility is improved for the transmitting side and the other side, and a smooth conversation can be achieved.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of a telephone according to the present invention.

FIG. 2 is a configuration diagram showing a schematic function of a DSP according to the present invention.

FIG. 3 is an external view of a telephone according to an embodiment of the present invention.

FIG. 4 is a schematic configuration diagram of a conventional portable telephone.

[Explanation of Codes] 1 Numeric keypad 2 Hook button 3 Transmitter / receiver 6 Speaker 8 Transmitter 9 Attenuator 15 Microphone 17 A / D converter 18 A / D converter 20 DSP 21 D / A converter 31a to 31m Delay element group 33a ˜33 m Linear prediction element group 35 Addition means 38 Transfer function calculation means 38 Judgment means

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Shinsuke Takada 1-7-12 Toranomon, Minato-ku, Tokyo Oki Electric Industry Co., Ltd.

Claims (1)

[Claims] 1. A microphone that receives an ambient sound and outputs it as an ambient sound signal, and a microphone that is arranged at a position different from the microphone and that receives the ambient sound or the ambient sound and the voice of a speaker and receives the sound. A transmitter that outputs as a signal, and a pseudo ambient sound signal is generated by the ambient sound signal from the microphone and a predetermined transfer function within a predetermined time with the start of a transmitting or receiving operation. The sound signal is compared with the sound reception signal due to the ambient sound from the transmitter, the transfer function is updated and set so that the difference between the two signals is minimized, and the transfer is performed after the predetermined time has elapsed. From a sound reception signal output from the device, a component of the pseudo ambient sound signal at that time is removed and output as a transmission signal, and an attenuator that outputs the transmission signal from the calculation unit to a speaker. , From the computing means Telephone, characterized in that it comprises sending a transmission signal to the other party, and a receiving unit which receives and outputs transmission signals from the mating side to the loudspeaker.