JP4826463B2 - Gateway device, control method thereof and program thereof - Google Patents

Description

  The present invention relates to a gateway device that relays between a public switched telephone network and an IP (Internet Protocol) telephone network.

There is a gateway device that relays between a public switched telephone network (PSTN network) and an IP (Internet Protocol) telephone network. When an audio signal is transmitted from the IP telephone network side to the public switched telephone network via the gateway device, a part of the audio signal is reflected by the hybrid circuit in the gateway device, and an echo is generated on the IP telephone network side. The problem arises.
For this reason, a communication apparatus having a function of removing echoes by an echo canceller has been realized (for example, Patent Document 1).
Japanese Patent Laid-Open No. 07-303069

  High response performance is required to quickly remove echoes with an echo canceller, and a high-load echo with an algorithm that can converge at high speed using a high-performance CPU (Central Processing Unit) or DSP (Digital Signal Processor). It is necessary to configure a canceller.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a gateway device capable of quickly removing echoes with a relatively low load.

To achieve the above object, a gateway device according to a first aspect of the present invention is a gateway device that relays a voice signal between an IP (Internet Protocol) network and a public switched telephone network, and the IP network The voice signal supplied via the transmission line is converted into a voice signal conforming to the communication protocol of the public telephone switched network, and is sent to the public switched telephone network. A voice signal transmitting / receiving means for converting the supplied voice signal into a voice signal conforming to the communication protocol of the IP network and transmitting the voice signal to the IP network via a reception line; and inputting the voice signal on the transmission line; Correction filter means for filtering and outputting on the reception line; A subtracting means for subtracting the output signal of the correction filter means from the audio signal on the reception line, a pseudo audio signal generating means for generating a pseudo audio signal, and a pseudo audio signal generated by the pseudo audio signal generating means The correction filter means so that the echo signal generated in the reception line and the output signal of the correction filter means substantially coincide with each other when the pseudo voice signal is reflected by the voice transmission / reception means. Correction filter setting means for setting the characteristics of the subtracting means, and echo cancellation means for following the change of the echo signal included in the output signal of the subtracting means and canceling the echo signal, and the correction filter setting means Call a public switched telephone network, stop the standby sound from the public switched network, After the stop, the setting processing of the characteristics of the correction filter means, characterized in that.
The correction filter setting means may determine the line type on the public switched telephone network side and stop the standby sound by a method suitable for the determined line type.
While the characteristic setting process of the correction filter unit is being performed, the pseudo audio signal may be input to the echo cancellation unit.

  A switching switch that switches between a sound signal and a pseudo sound signal and supplies the sound signal to the correction filter means and the sound signal transmitting / receiving means; and the correction filter setting means switches the changeover switch at a predetermined timing to switch the pseudo sound signal. To the correction filter means and the audio signal transmission / reception means to adjust the characteristics of the correction filter means, and after the adjustment is completed, the changeover switch is switched to convert the audio signal into the correction filter means and the audio signal transmission / reception means. At the predetermined timing, the output signal of the subtracting means is supplied to the correction filter means, and after the adjustment is completed, the supply is stopped, and the correction filter means supplies the output signal of the subtracting means. While there is, adjust its own characteristics so that the signal level of the output signal of the subtracting means is below a predetermined level, When the supply is exhausted, it may consist adaptive filter to maintain the regulated properties.

  The pseudo audio signal generating means may generate a pseudo audio signal having a frequency band substantially equal to a frequency band of a voice uttered by a human.

To achieve the above object, a control method of a gateway device according to a second aspect of the present invention, IP control method of a gateway device for relaying the voice signals between the (Internet Protocol) network and a public switched telephone network A voice signal supplied from the IP network via a transmission line is converted into a voice signal conforming to a communication protocol of the public switched telephone network and sent to the public switched telephone network; A voice signal transmission / reception step of converting a voice signal supplied from a public switched telephone network into a voice signal conforming to the communication protocol of the IP network and transmitting the voice signal to the IP network via a reception line; A correction filter that inputs audio signals, filters them, and outputs them A subtraction step that is arranged on the reception line and subtracts the output signal of the correction filter step from the audio signal on the reception line; a pseudo audio signal generation step that generates a pseudo audio signal; and the pseudo audio signal The pseudo audio signal generated in the generation step is supplied to the transmission line, and the echo signal generated in the reception line by the reflection of the pseudo audio signal in the audio transmission / reception step and the output signal of the correction filter step are substantially A correction filter setting step for setting the characteristics of the correction filter step to match, and an echo cancellation step for following the change in the echo signal included in the output signal of the subtraction step and canceling the echo signal , The correction filter setting step is performed on the public switched telephone network. And call, stop listening sound from the public switched telephone network, after stopping the waiting tone, the setting processing of the characteristics of the correction filter step, it is characterized.

In order to achieve the above object, a program according to a third aspect of the present invention is a program that causes a computer to realize a function of a gateway device that relays a voice signal between an IP (Internet Protocol) network and a public switched telephone network. The computer converts a voice signal supplied from the IP network via a transmission line into a voice signal conforming to a communication protocol of the public switched telephone network, and transmits the voice signal to the public switched telephone network. Voice signal transmitting / receiving means for transmitting, converting a voice signal supplied from the public switched telephone network to a voice signal conforming to a communication protocol of the IP network, and transmitting the voice signal to the IP network via a reception line; Input voice signal on talk line and filter Correction filter means for outputting the output signal, subtracting means arranged on the reception line and subtracting the output signal of the correction filter means from the sound signal on the reception line, and pseudo sound signal generation for generating a pseudo sound signal And a pseudo voice signal generated by the pseudo voice signal generating means is supplied to the transmission line, and an echo signal generated in the reception line when the pseudo voice signal is reflected by the voice transmitting / receiving means and the correction filter means Correction filter setting means for setting the characteristics of the correction filter means so that the output signals substantially match, and an echo that follows the change of the echo signal contained in the output signal of the subtraction means and cancels the echo signal canceling means, to function as the correction filter setting unit, a call to the public switched telephone network, the public The waiting sound from talking switched network stopped, after stopping the waiting tone, the setting processing of the characteristics of the correction filter means, characterized in that.

  According to the present invention, since at least part of the echo signal is removed by the correction filter means, the echo signal to be removed by the echo canceller means becomes relatively small, and the echo signal is quickly removed at a relatively low load. It becomes possible to do.

  Hereinafter, a communication system including a gateway device according to an embodiment of the present invention will be described with reference to the drawings.

As shown in FIG. 1, the communication system of this embodiment includes a gateway device (hereinafter referred to as GW) 1, a PSTN (Public Switched Telephone Networks) network 2, a telephone 3 (3 _{1 to} 3 _n ), an IP (Internet Protocol) network 4 and IP telephones 5 (5 _{1 to} 5 _n ).

  The GW 1 relays voice data transmitted and received between the PSTN network 2 and the IP network 4. As shown in FIG. 2, the GW 1 includes a storage unit 11, a control unit 12, and a communication control unit 13.

  The storage unit 11 includes a RAM (Random Access Memory), a ROM (Read Only Memory), and the like, and stores an operation program to be executed by the control unit 12.

  The control unit 12 includes a CPU (Central Processing Unit), a microprocessor, and the like, and controls the operation of the GW 1 by executing a predetermined operation program stored in the storage unit 11. For example, the control unit 12 controls the communication control unit 13 to perform processing for removing an echo generated by a voice signal transmitted from the IP network 4. This process will be described later.

The communication control unit 13 controls voice data communication between the IP network 4 and the PSTN network 2. As shown in FIG. 3, the communication control unit 13 includes a pseudo sound generator 21, switches SW1 and SW2, an echo canceller circuit 22, a correction filter 23, a hybrid circuit 24, and a D / A (Digital to Analog). A converter 25, an A / D (Analog to Digital) converter 26, amplifiers 27A and 27B, and subtraction circuits 28A and 28B are provided.
Note that the communication control unit 13 sequentially transmits information indicating the operation state to the control unit 12.

  The pseudo audio generator 21 generates a pseudo audio signal for setting the filter coefficient of the correction filter 23. The pseudo audio signal generated by the pseudo audio generator 21 is set so that the average audio signal uttered by humans matches the frequency band and frequency components.

The switches SW1 and SW2 are for switching the path of the audio signal.
The switch SW1 is a switch for switching the input signal of the hybrid circuit 24 between the voice signal generated by the pseudo voice generator 21 and the voice signal transmitted from the IP network 4.
The switch SW2 is a switch for switching whether or not to feed back the output signal of the subtraction circuit 28A to the correction filter 23.

The echo canceller circuit 22 receives the audio signal on the transmission path Sa and the audio signal on the reception path Sb, and obtains an echo signal (component) included in the signal flowing on the reception path Sb by a predetermined algorithm. An echo cancel signal corresponding to the obtained echo signal is output.
For example, the echo canceller circuit 22 is composed of an active filter or the like, and is an LMS (Least Mean Square) of a difference between a signal obtained by processing a voice signal on the transmission path Sa based on a predetermined algorithm and an output signal of the subtraction circuit 28B. The filter count is set so that is minimized, and the audio signal on the transmission path Sa is filtered and output.

  The correction filter 23 is an adaptive filter that filters and outputs an input audio signal with a filter characteristic corresponding to a filter coefficient. While the switch SW2 is closed, the correction filter 23 feeds back the output signal of the subtraction circuit 28A and corrects its own filter coefficient so that the LMS of the signal becomes minimum (below the reference level). Set. On the other hand, when the switch SW2 is opened, the feedback signal is not supplied to the correction filter 23, and the correction filter 23 maintains the filter coefficient.

When the hybrid circuit 24 transmits a voice signal from the IP telephone 5 to the telephone 3, the hybrid circuit 24 converts the voice signal transmitted from the IP network 4 through the transmission path Sa into a format compatible with the communication protocol of the PSTN network 2. The voice signal is converted and output to the PSTN network 2. Since the conversion of the voice signal is not complete, a part of the voice signal flows back to the transmission source IP network 4 through the reception path Sb and becomes an echo signal.
When the hybrid circuit 24 transmits a voice signal from the telephone 3 to the IP telephone 5, the hybrid circuit 24 converts the voice signal transmitted from the PSTN network 2 into a voice signal conforming to the communication protocol of the IP network 4, and receives the voice signal. The data is output to the IP network 4 via the route Sb.

The D / A converter 25 converts the audio signal transmitted from the IP network 4 or the digital audio signal transmitted from the pseudo audio generator 21 into an analog audio signal.
The A / D converter 26 converts the analog audio signal output from the hybrid circuit 24 into a digital audio signal.

The amplifier 27A amplifies the audio signal output from the D / A converter 25.
The amplifier 27B amplifies the audio signal output from the hybrid circuit 24 to the reception path Sb. Note that an echo signal is also output from the hybrid circuit 24 to the reception path Sb. Therefore, the amplifier 27B also amplifies this echo signal. The echo signal is amplified into a larger echo signal.

  The subtracting circuit 28A subtracts the audio signal filtered by the correction filter 23 from the audio signal converted into a digital format by the A / D converter 26. The output signal of the correction filter 23 substantially matches the echo signal generated by the pseudo audio signal. Therefore, most of the echo signal included in the audio signal on the reception path Sb is removed by this subtraction process.

  The subtraction circuit 28B subtracts the output signal of the echo canceller circuit 22 from the audio signal after passing through the subtraction circuit 28A. By this subtraction processing, the remaining echoes that could not be removed by the subtraction circuit 28A are also removed.

Returning to FIG. 1, the PSTN network 2 is a network network constituted by general subscriber telephone lines. A plurality of telephones 3 (3 _{1 to} 3 _n ) are connected to the PSTN network 2.
The IP network 4 is a network network such as the Internet, for example. A plurality of IP telephones 5 (5 _{1 to} 5 _n ) are connected to the IP network 4.

Subsequently, the operation of the GW 1 of the communication system according to the present embodiment will be described.
The telephone 3 and the IP telephone 5 communicate with each other via the GW 1. During this time, the GW 1 supplies the digital audio signal from the IP telephone 5 to the D / A converter 25 via the transmission path Sa. The D / A converter 25 converts the supplied digital audio signal into an analog audio signal and outputs the analog audio signal to the hybrid circuit 24 via the amplifier 27A.
The hybrid circuit 24 transmits the supplied analog audio signal to the telephone 3 via the PSTN network 2. At this time, due to impedance mismatch between the PSTN network 2 and the hybrid circuit 24, a part of the audio signal on the transmission path Sa becomes an echo signal (component) and leaks to the reception path Sb. Since the impedance of the PSTN network 2 varies depending on the line distance and the like, the degree of impedance mismatch with the hybrid circuit 24 that is a fixed circuit varies depending on which PSTN network 2 is connected to the hybrid circuit 24. For this reason, the level of the echo signal leaking to the reception path Sb also changes depending on the installation environment.

  On the other hand, the GW 1 receives the analog audio signal from the telephone 3 by the hybrid circuit 24, amplifies it by the amplifier 27 B, and supplies it to the A / D converter 26. At this time, the analog signal supplied to the A / D converter 26 includes a voice signal (original voice signal) from the PSTN network 2 and a voice signal not output to the PSTN network 2 by the hybrid circuit 24, that is, an echo component. Is an analog signal. Hereinafter, in order to facilitate understanding, a synthesized signal of the original audio signal and the echo signal is simply referred to as an audio signal.

The A / D converter 26 converts the analog audio signal into a digital audio signal and outputs it.
The subtraction circuit 28A subtracts the output signal of the correction filter 23 from this digital audio signal. The filter coefficient of the correction filter 23 is set so that a signal substantially equal to the echo signal included in the voice signal on the reception path Sb is generated from the voice signal on the transmission path Sa by a learning operation described later. . Therefore, most of the echo signal included in the audio signal passing through the reception path Sb is removed.

  The output signal of the subtraction circuit 28A is supplied to the subtraction circuit 28B. The subtraction circuit 28B subtracts the output signal of the echo canceller circuit 22 from this output signal. The echo canceller circuit 22 is supplied with the audio signal on the transmission path Sa and is always fed back with the output signal of the subtracting circuit 28B. The echo canceller circuit 22 is a receiving circuit similar to a normal echo canceller circuit. A signal substantially equal to the echo signal included (remaining) in the audio signal on Sb is output. For this reason, the subtraction circuit 28B substantially cancels and outputs the echo signal included (remaining) in the audio signal from the subtraction circuit 28A. This voice signal, that is, a high-quality voice signal from which the echo signal has been canceled is transmitted to the IP telephone 5 via the IP network 4.

In order to perform such echo cancellation processing efficiently and appropriately, the filter coefficient of the correction filter 23 is appropriately set so that the correction filter 23 can generate and output a signal approximate to the echo signal. There is a need to.
Therefore, the control unit 12 performs a correction filter setting process for setting an appropriate filter coefficient in the correction filter 23 at a predetermined timing.
The correction filter setting process will be described in detail with reference to FIG.
This correction filter setting process is started when the control unit 12 receives information indicating that the power of the GW 1 has been turned on or instruction information for performing an echo cancellation learning process.

  First, when the correction filter setting process is started, the control unit 12 off-hooks the line and determines the type of line (push line or dial line) on the PSTN network 2 side connected to itself (step S11). For example, the control unit 12 transmits a DTMF (Dial Tone Multi Frequency) signal to detect a dial tone. If a dial tone is detected, the control unit 12 determines a dial line, and if a dial tone is not detected, determines a push line.

  Next, the control unit 12 dials a fictitious terminal device connected to the PSTN network 2 in a method that matches the type of the line determined in step S11, and waits for an off-hook standby sound (a so-called dial tone (DT). )) Is not generated (step S12). This process is for preventing DT from occurring, and it is not necessary to make a call.

  Subsequently, the control unit 12 switches the switch SW <b> 1 so that the pseudo sound signal generated by the pseudo sound generator 21 is input to the hybrid circuit 24, the correction filter 23, and the echo canceller circuit 22. In addition, the control unit 12 closes the switch SW2 so that the audio signal output from the subtraction circuit 28A is fed back to the correction filter 23 (step S13).

  Subsequently, the control unit 12 controls the pseudo audio generator 21 to output a pseudo audio signal, and corrects the echo signal generated from the hybrid circuit 24 by the pseudo audio signal so that it can be canceled by the subtracting circuit 28A. The filter coefficient of the filter 23 is set (step S14). Specifically, the correction filter 23 sets its own filter coefficient so that the level or LMS of the output signal of the subtraction circuit 28A fed back to the correction filter 23 is minimized (below the reference level).

When a predetermined time (for example, 20 s) sufficient to set the filter coefficient of the correction filter 23 has elapsed (step S15; Yes), the control unit 12 determines that the audio signal from the IP network 4 is the hybrid circuit 24, The switch SW1 is switched so as to be input to the correction filter 23 and the echo canceller circuit 22. Further, the control unit 12 opens the switch SW2 so that the output signal of the subtraction circuit 28A is not fed back to the correction filter 23 (step S16). By this processing, the output signal of the subtracting circuit 28A is not fed back to the correction filter 23, so that the filter coefficient of the correction filter 23 is fixed.
This is the end of the correction filter setting process.

  With this correction filter setting process, the filter coefficient of the correction filter 23 is set to a value most suitable for canceling the echo signal generated from the hybrid circuit 24 by the sound signal output from the pseudo sound generator 21. The audio signal output from the pseudo audio generator 21 is set to have a frequency band that is substantially equal to the frequency band of the audio signal uttered by a human. For this reason, it can be said that the filter coefficient of the correction filter 23 is set to an optimum value for canceling an echo generated by a voice generated by a human.

  As described above, according to the GW 1 having the above-described configuration, the echo signal included in the audio signal on the reception path Sb is canceled by the correction filter 23 learned in advance using the pseudo audio signal. Therefore, the echo signal that needs to be removed by the echo canceller circuit 22 and the subtractor circuit 28B becomes smaller. Therefore, even if the computation capability and speed of the echo canceller circuit 22 are relatively small, it is possible to respond quickly to changes in the echo signal. Further, the correction filter 23 can set its filter coefficient over time in the absence of the audio signal from the PSTN network 2. Accordingly, the processing capability and arithmetic operation capability of the correction filter 23 are not required to be large, and so-called response characteristics may be low. Therefore, it is possible to obtain an echo canceller capable of promptly responding to fluctuations in the echo signal, although the whole can be configured using a computing element having a relatively low computing capability.

  In addition, this invention is not limited to the said embodiment, A various deformation | transformation and application are possible.

  For example, in the above-described embodiment, a circuit composed of discrete components is shown as the configuration of the communication control unit 13. However, for example, the correction filter 23, the echo canceller circuit 22, the pseudo sound generator 21, etc. You may comprise as a part of functions, such as.

  In the above embodiment, in order to cancel the echo signal, the output signal of the correction filter 23 is subtracted from the audio signal by the subtraction circuit 28A. However, the polarity of the output signal of the correction filter 23 is inverted and added to the audio signal. Etc. In this specification, the configuration for obtaining (cancelling) the difference between these two echo signals is called a subtracting circuit, but the specific circuit configuration itself is arbitrary.

The timing at which the correction filter 23 performs learning (correction filter setting processing) is not limited to the above embodiment, and is arbitrary, and may be periodically executed at, for example, midnight.
Further, the number of correction filters 23 is not limited to one and may be plural. The number of echo canceller circuits 22 is also arbitrary.

  The circuit configuration illustrated in FIG. 3, the flowchart illustrated in FIG. 4, and the like are examples, and the configuration and functions thereof are arbitrary as long as the same results as the contents disclosed in the above embodiments can be obtained.

  Further, the function performed by the control unit 12 of the GW 1 according to the above embodiment can be realized by a general-purpose microprocessor or DSP. In this case, you may make a processor run the program for implement | achieving said each process.

It is the figure which showed the structure of the communication system which concerns on one embodiment of this invention. It is the block diagram which showed the structure of the gateway apparatus contained in the communication system shown in FIG. It is the figure which showed the structure of the communication control part of the gateway apparatus shown in FIG. 6 is a flowchart for explaining an operation of a correction filter setting process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 13 Communication control part 21 Pseudo sound generator 22 Echo canceller circuit 23 Correction filter 24 Hybrid circuit 25 D / A converter 26 A / D converter 27A, 27B Amplifier 28A, 28B Subtraction circuit 2 PSTN network 4 IP network Sa Transmission path Sb listening path

Claims (7)

A gateway device that relays audio signals between an IP (Internet Protocol) network and a public switched telephone network,
A voice signal supplied from the IP network via a transmission line is converted into a voice signal conforming to a communication protocol of the public telephone switched circuit network, and is sent to the public telephone switched circuit network. Voice signal transmitting / receiving means for converting a voice signal supplied from a line network into a voice signal conforming to a communication protocol of the IP network and transmitting the voice signal to the IP network via a receiving line;
Correction filter means for inputting a voice signal on the transmission line, filtering and outputting, and
A subtracting means arranged on the reception line and subtracting the output signal of the correction filter means from a voice signal on the reception line;
Pseudo audio signal generating means for generating a pseudo audio signal;
The pseudo voice signal generated by the pseudo voice signal generation means is supplied to the transmission line, and the pseudo voice signal is reflected by the voice transmission / reception means, and the echo signal generated in the reception line and the output signal of the correction filter means Correction filter setting means for setting the characteristics of the correction filter means so that substantially matches,
An echo canceling means for following the change of the echo signal contained in the output signal of the subtracting means and canceling the echo signal;
Equipped with a,
The correction filter setting means includes
Making a call to the public switched telephone network, stopping the standby sound from the public switched telephone network,
After the standby sound is stopped, the correction filter means characteristics are set.
A gateway device characterized by that.   The correction filter setting means determines the line type on the public switched telephone line network side, and stops the standby sound in a method suitable for the determined line type.
  The gateway device according to claim 1. While the process for setting the characteristics of the correction filter means is performed, the pseudo audio signal is input to the echo cancellation means.
  The gateway device according to claim 1 or 2, wherein A changeover switch for switching between an audio signal and a pseudo audio signal to supply to the correction filter means and the audio signal transmitting / receiving means
The correction filter setting means includes
At a predetermined timing, the selector switch is switched to supply a pseudo audio signal to the correction filter means and the audio signal transmitting / receiving means to adjust the characteristics of the correction filter means, and after completion of the adjustment, the selector switch is switched. Supplying the audio signal to the correction filter means and the audio signal transmitting / receiving means,
At the predetermined timing, the output signal of the subtracting means is supplied to the correction filter means, and after the adjustment is finished, the supply is stopped,
The correction filter means adjusts its characteristics so that the signal level of the output signal of the subtracting means is below a predetermined level while the output signal of the subtracting means is supplied, and when the supply is lost Consisting of an adaptive filter that maintains the adjusted characteristics,
The gateway apparatus according to any one of claims 1 to 3, wherein The pseudo audio signal generating means generates a pseudo audio signal having a frequency band substantially equal to a frequency band of a voice uttered by a human.
The gateway apparatus according to any one of claims 1 to 4 , wherein the gateway apparatus is characterized in that: A control method for a gateway device that relays an audio signal between an IP (Internet Protocol) network and a public switched telephone network,
A voice signal supplied from the IP network via a transmission line is converted into a voice signal conforming to a communication protocol of the public telephone switched circuit network, and is sent to the public telephone switched circuit network. A voice signal transmitting / receiving step of converting a voice signal supplied from a line network into a voice signal conforming to a communication protocol of the IP network and transmitting the voice signal to the IP network via a reception line;
A correction filter step for inputting, filtering and outputting a voice signal on the transmission line;
A subtracting step disposed on the reception line and subtracting the output signal of the correction filter step from the audio signal on the reception line;
A pseudo audio signal generating step for generating a pseudo audio signal;
The pseudo audio signal generated in the pseudo audio signal generation step is supplied to the transmission line, and the pseudo audio signal is reflected in the audio transmission / reception step, and the echo signal generated in the reception line and the output signal of the correction filter step A correction filter setting step for setting the characteristic of the correction filter step so that substantially matches
An echo canceling step for following the change of the echo signal contained in the output signal of the subtraction step and canceling the echo signal;
Equipped with a,
The correction filter setting step includes:
Making a call to the public switched telephone network, stopping the standby sound from the public switched telephone network,
After the standby sound is stopped, the correction filter step characteristic setting process is performed.
A control method for a gateway device. A program that causes a computer to realize a function of a gateway device that relays an audio signal between an IP (Internet Protocol) network and a public switched telephone network,
A voice signal supplied from the IP network via a transmission line is converted into a voice signal conforming to a communication protocol of the public telephone switched circuit network, and is sent to the public telephone switched circuit network. Voice signal transmitting / receiving means for converting a voice signal supplied from a line network into a voice signal conforming to a communication protocol of the IP network and transmitting the voice signal to the IP network via a receiving line;
Correction filter means for inputting a voice signal on the transmission line, filtering and outputting, and
A subtracting means arranged on the reception line and subtracting the output signal of the correction filter means from a voice signal on the reception line;
Pseudo audio signal generating means for generating a pseudo audio signal;
The pseudo voice signal generated by the pseudo voice signal generation means is supplied to the transmission line, and the pseudo voice signal is reflected by the voice transmission / reception means, and the echo signal generated in the reception line and the output signal of the correction filter means Correction filter setting means for setting the characteristics of the correction filter means so that substantially matches,
An echo canceling means for following the change of the echo signal contained in the output signal of the subtracting means and canceling the echo signal;
To function as,
The correction filter setting means includes
Making a call to the public switched telephone network, stopping the standby sound from the public switched telephone network,
After the standby sound is stopped, the correction filter means characteristics are set.
A program characterized by that.

Images