Intelligent PSTN Telephone Voice Channel Switching Noise Cancellation Method

Abstract

The present invention relates to the technical field of intelligent PSTN telephone manufacturing, specifically an intelligent PSTN telephone particularly suitable for being equipped with an intelligent operating system, and an intelligent PSTN telephone voice channel switching noise cancellation method capable of eliminating noise during the switching process of multiple voice channels, thereby improving the user experience. It is characterized in that a voice channel switching unit is further provided. The microphone unit and the hands-free unit are respectively connected to the voice channel of the PSTN line through the voice channel switching unit. A mute noise reduction and echo cancellation unit is further provided between the voice channel switching unit and the voice channel of the PSTN line. The control end of the mute noise reduction and echo cancellation unit is connected to the central control unit. Compared with the prior art, through the switching control of multiple audio channels in the intelligent PSTN telephone by the central control unit, different control processes are used in different states to eliminate line switching noise, and the user's perception of the electronic switch switching noise can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent PSTN telephone manufacturing, and specifically to an intelligent PSTN telephone that is particularly suitable for being equipped with an intelligent operating system, capable of eliminating noise during the switching of multiple voice channels, and further improving the intelligent PSTN telephone voice channel switching noise cancellation method for the use experience. Background Art

[0002] PSTN is a telephone switching network based on analog technology, and the channels are connected via several telephone switches. Traditional PSTN telephones are not equipped with an intelligent operating system, so they can only provide two voice output modes: hands-free call and microphone call. Since the voice channels of the hands-free and the microphone use a plug-in switch for switching, a clicking pulse sound of the switch can be heard during the switching, whether it is for sending or receiving, which affects the call quality.

[0003] In order to further improve the device functions, existing intelligent PSTN telephones can be equipped with an intelligent operating system based on Android or Linux. Telephones equipped with an intelligent operating system can not only achieve voice output in two modes: hands-free and microphone, but also have a media voice channel. At present, there are pulse noises generated by the switching of electronic switches during the switching between the media voice channel, the hands-free voice channel, and the microphone voice channel. Summary of the Invention

[0004] The present invention aims at the disadvantages and deficiencies existing in the prior art, and proposes an intelligent PSTN telephone voice channel switching noise cancellation method that is particularly suitable for an intelligent PSTN telephone equipped with an intelligent operating system, capable of eliminating noise during the switching of multiple voice channels, and further improving the use experience.

[0005] The present invention is achieved through the following measures:

[0006] An intelligent PSTN telephone voice channel switching noise cancellation method, in the intelligent PSTN telephone, there is a central control unit with a media audio channel, a microphone unit with a microphone audio channel, a hands-free unit with a hands-free audio channel, and a PSTN line voice channel. It is characterized in that there is also a voice channel switching unit. The microphone unit and the hands-free unit are respectively connected to the PSTN line voice channel through the voice channel switching unit. There is also a mute noise reduction and echo cancellation unit between the voice channel switching unit and the PSTN line voice channel. The control end of the mute noise reduction and echo cancellation unit is connected to the central control unit. The intelligent PSTN telephone voice channel switching noise cancellation method includes the following content:

[0007] When the handset call mode is executed, the handset unit goes off-hook, and the central control unit controls the action of the electronic switch in the voice channel switching unit, connecting the media audio channel to the handset audio channel. The media audio channel plays sound through the handset speaker, and the APP dials to emit an analog key tone, which the user can hear on the handset. After that, the central control unit controls the mute noise reduction and echo cancellation unit to mute the external line, starts sending DTMF numbers to the PSTN external line, then turns off the external line mute to restore the external line voice, and waits for the external line to be connected to start the call;

[0008] When the hands-free call mode is executed, the central control unit controls the action of the electronic switch in the voice channel switching unit, connecting the media audio channel to the hands-free audio channel. The APP plays sound through the hands-free speaker, and the APP dials to emit an analog key tone, which the user can hear on the hands-free speaker. After that, the central control unit controls the mute noise reduction and echo cancellation unit to mute the external line, starts sending DTMF numbers to the PSTN external line, and then the central control unit controls the mute noise reduction and echo cancellation unit to turn off the external line mute to restore the external line voice, and waits for the external line to be connected to start the call;

[0009] When the handset-to-hands-free call mode is executed, in the handset call state, click the hands-free call button on the APP call interface in the central control unit, and set the volume of the media handset channel in the media audio channel to 0. The mute noise reduction and echo cancellation unit controls the external line to be muted, and the pulse noise generated by the switching of the electronic switch in the voice channel switching unit will not be transmitted to the PSTN external line, so the other end of the call will not hear the switching noise. The electronic switch in the voice channel switching unit acts, first disconnecting the handset audio channel from the hands-free audio channel, and then connecting the hands-free voice channel to the PSTN line voice channel through the mute noise reduction and echo cancellation unit. Through the central control unit, the mute noise reduction and echo cancellation unit sequentially executes turning off the external line mute and restoring the external line voice, and the media handset channel resumes the volume to continue the call;

[0010] When the hands-free-to-handset call mode is executed, in the hands-free call state, the handset unit goes off-hook, and the mute noise reduction and echo cancellation unit controls the PSTN line voice channel of the external line to be muted. The pulse noise generated by the switching of the electronic switch in the voice switching control unit will not be transmitted to the PSTN external line, so the other end of the call will not hear the switching noise: The electronic switch in the voice channel switching unit acts, first disconnecting the handset audio channel from the PSTN line audio channel connected through the mute noise reduction and echo cancellation unit, then connecting the external line to the handset, and through the central control unit, controlling the mute noise reduction and echo cancellation unit to turn off the external line mute and restore the external line voice, and the media handset channel resumes the volume to continue the call.

[0011] When the present invention performs a second dialing in the handset call state, initially in the handset call state, click the numeric keypad on the APP call interface of the central control unit to send DTMF numbers, and the mute, noise reduction, and echo cancellation unit controls the external line to be muted, ensuring that the pulse noise generated by the electronic switch switching will not be transmitted to the PSTN external line. Specifically: the media audio channel in the central control unit sets the volume to 0, ensuring that the noise generated by the electronic switch switching in the voice channel switching unit cannot be heard on the local handset. The electronic switch in the voice channel switching unit operates to connect the handset audio channel to the media audio channel, and the media audio channel restores the volume. The media audio channel can play sounds through the handset speaker, and the APP dialing emits simulated key tones that the user can hear on the handset. After the PSTN number sending is completed, the media audio channel sets the volume to 0, the electronic switch in the voice channel switching unit operates, and connects the handset audio channel to the PSTN line audio channel connected through the mute, noise reduction, and echo cancellation unit. The media audio channel restores the volume, turns off the mute state of the mute, noise reduction, and echo cancellation unit, and the call can continue.

[0012] When the present invention performs a second dialing in the hands-free call state, initially in the hands-free call state, click the numeric keypad on the APP call interface of the central control unit to send DTMF numbers, and the mute, noise reduction, and echo cancellation unit controls the external line to be muted, ensuring that the pulse noise generated by the electronic switch switching will not be transmitted to the PSTN external line. Specifically: the media audio channel sets the volume to 0, ensuring that the noise generated by the electronic switch switching cannot be heard on the local handset. The electronic switch of the voice channel switching unit operates to connect the hands-free audio channel to the media voice channel, and the media voice channel restores the volume. The media audio channel can play sounds through the hands-free speaker, and the APP dialing emits simulated key tones that the user can hear on the hands-free speaker. After the PSTN number sending is completed, the media voice channel sets the volume to 0, the electronic switch in the voice channel switching unit operates again to connect the hands-free audio channel to the PSTN line voice channel connected through the mute, noise reduction, and echo cancellation unit. The media audio channel restores the volume, turns off the mute of the mute, noise reduction, and echo cancellation unit, and the call can continue.

[0013] In the present invention, the mute and noise reduction and echo cancellation unit is provided with a volume control module for implementing mute processing, an echo cancellation module for implementing echo cancellation processing, and a noise reduction module for implementing noise reduction processing of the voice path. The noise reduction module adopts a dual-threshold and dual-judgment noise reduction method, which has an enhanced noise reduction effect on the sound frequency of the "click" noise generated when the sound channel is switched on the line. Specifically, let a certain line signal be s, the signal after calculation by the noise reduction algorithm be r, s(k) be the voice signal of the k-th frame, k represents the subscript of the k-th frame, its short-time zero-crossing rate is z(k) (the ratio of the number of times the signal passes through the 0 point after k frames), the short-time power is e(k) (the signal power of the audio of the k-th frame), the frame shift is Ts, and then let the zero-crossing rate threshold be zTol and the power threshold be eTol.

[0014] Let the enhanced speech after noise suppression be r(k) = s(k), and the background noise suppression criterion is:

[0015] When e(k) < eTol and z(k) > zTol, then it is judged that s(k) is a noise signal, and the signal r(k) after noise reduction is set to 0, so the noise signal is eliminated;

[0016] When the continuous non-zero segment (from the k1-th frame to the k2-th frame) r(k1:k2)≠0 in the signal r after noise reduction, and the length (k2 - k1)Ts < 20ms, then let r(k1:k2)=0, further eliminating the noise in a short time, and a part of the "click" noise generated by the switching on the line is included in it and can be eliminated here;

[0017] After the above dual-threshold and dual-discrimination, r(k) will no longer contain non-speech interference and transient interference, that is, the "click" noise generated during the switching on the line. eTol is usually taken as 0.05 times the maximum value of s, and zTol is usually taken as 75% of the frame length. In the algorithm, the values of zTol and eTol are eTol = 0.05 and zTol = 50% respectively.

[0018] Compared with the prior art, the present invention controls the switching of multiple audio channels in the intelligent PSTN telephone through the central control unit, and uses different control processes in different states to eliminate the line switching noise, which can reduce the user's perception of the electronic switch switching noise. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Attached Figure 1 is the structural block diagram of the present invention.

[0020] Attached Figure 2 is the flow chart of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0021] The present invention will be further described below in conjunction with the drawings and embodiments.

[0022] Example 1:

[0023] As shown in the Figure 1 accompanying figure, this example provides a noise cancellation method for switching the sound channel of an intelligent PSTN phone. The intelligent PSTN phone described in this example is provided with:

[0024] 10 Central control unit, the unit where the core CPU is located, which completes the control of other units;

[0025] 20 Sound channel switching control unit, which includes controlling the connection between the hands-free channel and the PSTN line voice channel, controlling the connection between the handset channel and the PSTN line voice channel, and controlling the disconnection from the PSTN line voice channel and only using the media sound channel;

[0026] 30 Mute noise reduction and echo cancellation unit, which has the functions of controlling volume, reducing noise and canceling echo; 40 PSTN line voice channel, which connects to the PSTN line of the telephone;

[0027] 50 Microphone unit, which includes a microphone speaker and a microphone MIC;

[0028] 60 Hands-free unit includes a hands-free speaker and a hands-free MIC;

[0029] The 10 central control unit described in this example, that is, the unit where the core CPU is located, uses the Android or Linux operating system, which has display function, touch function, GPIO control function, echo cancellation, noise reduction function through noise reduction algorithm, volume control and mute control functions through volume control IC, and runs APP to support PSTN function, etc.;

[0030] As shown in the Figure 1As shown in the figure, the voice channel switching control unit described in this example consists of 4 groups of electronic switches. Each electronic switch is a 1-to-2 switch, that is, one input and two outputs. The high / low control input signal of a GPIO is used to control which output channel is connected. Among them, electronic switch 1 (201) controls the connection and disconnection of the mic in the microphone voice channel, the mic in the media microphone channel, and the mic in the PSTN line voice channel. Electronic switch 2 (202) controls the connection and disconnection of the speaker in the microphone voice channel, the speaker in the media microphone channel, and the speaker in the PSTN line voice channel. That is, electronic switch 1 (201) and electronic switch 2 (202) manage the connection and disconnection between audio channel 3 and audio channel 1 or audio channel 4. Electronic switch 3 (203) controls the connection and disconnection of the mic in the hands-free voice channel, the mic in the media hands-free channel, and the mic in the PSTN line voice channel. Electronic switch 4 (204) controls the connection and disconnection of the speaker in the hands-free voice channel, the speaker in the media hands-free channel, and the speaker in the PSTN line voice channel. That is, electronic switch 3 (203) and electronic switch 4 (204) manage the connection and disconnection between audio channel 2 and audio channel 0 or audio channel 4. When the voice channel is switched, pulses will be generated when the electronic switches act, and the "click" noise is introduced here.

[0031] The 30 mute, noise reduction, and echo cancellation unit described in this example has a mute control, a noise reduction function, and an echo cancellation function. The volume control and mute functions are controlled by the 10 central control unit through the I2C interface to control the 302 volume control IC. When the mute, noise reduction, and echo cancellation unit is muted, it will prevent the noise generated in the 40 PSTN voice channel unit when the electronic switch is switched, but it cannot eliminate the "click" noise heard in the 50 microphone unit or the 60 hands-free unit. Therefore, it is also necessary to run the 301 noise reduction algorithm through the 10 central control unit to eliminate the line noise, and then use the echo cancellation IC to eliminate the line echo and the unique control sequence of each unit in terms of software to implement the method of eliminating noise. The 301 noise reduction algorithm uses a dual-threshold and dual-judgment method, which has a strengthening noise reduction effect on the sound frequency of the "click" noise generated during switching on the line. The specific noise reduction algorithm is as follows:

[0032] Let a certain line signal be s, and the signal after being calculated by the noise reduction algorithm be r. s(k) is the voice signal of the k-th frame, k represents the subscript of the k-th frame, its short-time zero-crossing rate is z(k) (the ratio of the number of times the signal passes through the 0 point after k frames), and the short-time power is e(k) (the signal power of the audio of the k-th frame). Take the frame shift as Ts, and then set the zero-crossing rate threshold as zTol and the power threshold as eTol;

[0033] Let the enhanced voice after noise suppression be r(k) = s(k), and the background noise suppression criterion is:

[0034] When e(k) < eTol and z(k) > zTol, then s(k) is determined to be a noise signal, and the denoised signal r(k) is set to 0, thus eliminating the noise signal.

[0035] When there is a continuous non - zero segment in the denoised signal r (from frame k1 to k2), r(k1:k2)≠0, and the length (k2 - k1)Ts < 20ms, then let r(k1:k2)=0

[0036] Further eliminate the noise in a short period of time. A part of the "click" noise generated by the switching on the line is included in it and can be eliminated here.

[0037] After the above double - threshold and double - discrimination, r(k) will no longer contain non - speech interference and transient interference, that is, the "click" noise generated during the switching on the line. eTol is usually taken as 0.05 times the maximum value of s, and zTol is usually taken as 75% of the frame length. Through long - term debugging and verification of the values of zTol and eTol in the algorithm, we obtain the most suitable values, which are eTol = 0.05 and zTol = 50% respectively.

[0038] The 40 PSTN voice channel unit described in this example is a component connecting the audio channel 4 to the PSTN external line; the 50 microphone unit, that is, the handset of the telephone, includes a MIC and a speaker, as well as a coiled cord connecting to the audio channel 1; the 60 hands - free unit includes a hands - free MIC and a hands - free speaker.

[0039] As shown in the appendix Figure 2 The specific control process of this example is as follows:

[0040] When the handset call mode is executed: The 50 microphone unit goes off - hook. By controlling the electronic switch 1 and the electronic switch 2, the media audio channel is connected to the microphone voice channel, that is, the audio channel 3 is connected to the audio channel 1. The media audio channel can play sound through the microphone speaker. The APP dials to send out an analog key tone, which the user can hear on the handset. The control of the mute, noise reduction, and echo cancellation unit mutes the external line, starts to send DTMF numbers to the PSTN external line, then closes the external line mute to restore the external line voice, and waits for the external line to be connected to start the call.

[0041] When the hands - free call mode is executed: By controlling the electronic switch 3 and the electronic switch 4, the media audio channel is connected to the hands - free voice channel, the audio channel 2 is connected to the audio channel 0, and the 60 hands - free unit audio channel is connected to the media audio channel. The APP can play sound through the hands - free speaker. The APP dials to send out an analog key tone, which the user can hear on the hands - free speaker. The control of the mute, noise reduction, and echo cancellation unit mutes the external line, starts to send DTMF numbers to the PSTN external line, then controls the 30 mute, noise reduction, and echo cancellation unit to close the external line mute to restore the external line voice, and waits for the external line to be connected to start the call.

[0042] When the microphone cut-to-hands-free call mode is executed and the initial state is the microphone call state, click the hands-free call button on the central control unit APP call interface. The media microphone channel sets the volume to 0, and the 30 silent noise reduction and echo cancellation unit controls the external line to be muted. The pulse noise switched by the electronic switch will not be transmitted to the PSTN external line, and the other end of the call will not hear the switching noise. Control electronic switch 1 and electronic switch 2 to disconnect audio channel 3 and audio channel 2. Control electronic switch 3 and electronic switch 4 to connect audio channel 2 and audio channel 4. Then control the 30 silent noise reduction and echo cancellation unit to turn off the external line mute and restore the external line voice. The media microphone channel restores the volume and continues the call;

[0043] When the hands-free cut-to-microphone call mode is executed and the initial state is the hands-free call state, pick up the microphone unit at 50. The 30 silent noise reduction and echo cancellation unit controls the external line to be muted. The pulse noise switched by the electronic switch will not be transmitted to the PSTN external line, and the other end of the call will not hear the switching noise. Control electronic switch 3 and electronic switch 4 to disconnect audio channel 2 and audio channel 4. Then control electronic switch 1 and electronic switch 2 to connect audio channel 1 and audio channel 4, connect the external line and the microphone. Then control the 30 silent noise reduction and echo cancellation unit to turn off the external line mute and restore the external line voice. The media microphone channel restores the volume and continues the call;

[0044] When the second dialing is performed in the microphone call state and the initial state is the microphone call state, click the numeric keypad on the central control unit APP call interface to send the DTMF number at 10. The 30 silent noise reduction and echo cancellation unit controls the external line to be muted to ensure that the pulse noise switched by the electronic switch will not be transmitted to the PSTN external line. The media sound channel sets the volume to 0 to ensure that the switching noise of the electronic switch cannot be heard on the local microphone. Control electronic switch 1 and electronic switch 2 to connect the microphone voice channel and the media voice channel, that is, connect audio channel 3 and audio channel 1. The media sound channel restores the volume. The media audio channel can play the sound through the microphone speaker. The APP dials and emits an analog key tone, which the user can hear on the microphone. After the PSTN number sending is completed, the media sound channel sets the volume to 0. Control electronic switch 1 and electronic switch 2 to connect audio channel 3 and audio channel 4. The media sound channel restores the volume. Turn off the 30 silent noise reduction and echo cancellation unit mute and continue the call;

[0045] When performing a second dialing in the hands-free call state, initially in the hands-free call state, click the numeric keypad on the APP call interface of the central control unit to send DTMF numbers. The 30 silent noise reduction and echo cancellation unit controls the external line to be silent, ensuring that the pulse noise caused by the electronic switch switching will not be transmitted to the PSTN external line. The media sound channel sets the volume to 0, ensuring that the noise caused by the electronic switch switching cannot be heard on the local microphone. Control electronic switch 3 and electronic switch 4 to connect the hands-free voice channel to the media voice channel, that is, connect audio channel 2 to audio channel 0. The media sound channel restores the volume, and the media audio channel can play sound through the hands-free speaker. The APP dialing emits analog keypad tones, which the user can hear on the hands-free speaker. After the PSTN number sending is completed, the media sound channel sets the volume to 0, controls electronic switch 3 and electronic switch 4, connects audio channel 2 to audio channel 4, the media sound channel restores the volume, and turns off the silence of the 30 silent noise reduction and echo cancellation unit, and the call can continue.

[0046] Compared with the prior art, the present invention controls the switching of multiple audio channels in the intelligent PSTN phone through the central control unit, and uses different control processes in different states to eliminate the line switching noise, which can reduce the user's perception of the electronic switch switching noise.

Claims

1. A noise cancellation method for switching voice channels of an intelligent PSTN phone. In the intelligent PSTN phone, there is a central control unit with a media audio channel, a microphone unit with a microphone audio channel, a hands-free unit with a hands-free audio channel, and a PSTN line voice channel. It is characterized in that, There is also a voice channel switching unit. The microphone unit and the hands-free unit are respectively connected to the voice channel of the PSTN line through the voice channel switching unit. There is also a mute noise reduction and echo cancellation unit between the voice channel switching unit and the voice channel of the PSTN line. The control end of the mute noise reduction and echo cancellation unit is connected to the central control unit. The intelligent PSTN phone voice channel switching noise reduction method includes the following content: When the microphone call mode is executed, the microphone unit goes off-hook. The central control unit controls the electronic switch in the voice channel switching unit to act, so that the media audio channel is connected to the microphone audio channel. The media audio channel plays sound through the microphone speaker, and the APP dials to send out analog key tones, which can be heard by the user on the microphone. Thereafter, the central control unit controls the mute noise reduction and echo cancellation unit to mute the external line, starts to send DTMF numbers to the PSTN external line, then closes the external line mute and restores the external line voice, and waits for the external line to be connected to start the call; When the hands-free call mode is executed, the central control unit controls the electronic switch in the voice channel switching unit to act, and the media audio channel is connected to the hands-free audio channel. The APP plays sound through the hands-free speaker, and the APP dials to send out analog key tones, which can be heard by the user on the hands-free speaker. Thereafter, the central control unit controls the mute noise reduction and echo cancellation unit to mute the external line, starts to send DTMF numbers to the PSTN external line, and then the central control unit controls the mute noise reduction and echo cancellation unit to close the external line mute and restore the external line voice, and waits for the external line to be connected to start the call; When the microphone-to-hands-free call mode is executed, in the microphone call state, click the hands-free call button on the APP call interface in the central control unit, and set the volume of the media microphone channel in the media audio channel to 0. The mute noise reduction and echo cancellation unit controls the external line to be muted, and the pulse noise generated by the switching of the electronic switch in the voice channel switching unit will not be transmitted to the PSTN external line, and the other end of the call will not hear the switching noise. The electronic switch in the voice channel switching unit acts, first disconnects the microphone audio channel and the hands-free audio channel, and then makes the hands-free voice channel connected to the PSTN line voice channel through the mute noise reduction and echo cancellation unit. Through the central control unit, the mute noise reduction and echo cancellation unit sequentially executes closing the external line mute and restoring the external line voice, and the media microphone channel restores the volume to continue the call; When the hands-free-to-microphone call mode is executed, in the hands-free call state, the microphone unit goes off-hook. The mute noise reduction and echo cancellation unit controls the voice channel of the PSTN line of the external line to be muted, and the pulse noise generated by the switching of the electronic switch in the voice switching control unit will not be transmitted to the PSTN external line, and the other end of the call will not hear the switching noise: The electronic switch in the voice channel switching unit acts, first disconnects the hands-free audio channel and the PSTN line audio channel connected through the mute noise reduction and echo cancellation unit, then connects the external line to the microphone, and controls the mute noise reduction and echo cancellation unit to close the external line mute and restore the external line voice through the central control unit, and the media microphone channel restores the volume to continue the call.

2. The intelligent PSTN telephone voice channel switching noise cancellation method according to claim 1, wherein When performing a second dial during a handset call state, initially in the handset call state, click the numeric keypad on the central control unit APP call interface to send DTMF numbers. The mute noise reduction and echo cancellation unit controls the external line to be muted, ensuring that the pulse noise generated by the electronic switch switching will not be transmitted to the PSTN external line. Specifically: The media audio channel in the central control unit sets the volume to 0, ensuring that the noise generated by the electronic switch switching in the voice channel switching unit cannot be heard on the local handset. The electronic switch in the voice channel switching unit operates to connect the handset audio channel to the media audio channel. The media audio channel restores the volume, and the media audio channel plays the sound through the handset speaker. The APP dials to emit an analog key tone, which the user can hear on the handset. After the PSTN number sending is completed, the media audio channel sets the volume to 0, the electronic switch in the voice channel switching unit operates, and connects the handset audio channel to the PSTN line audio channel connected through the mute noise reduction and echo cancellation unit. The media audio channel restores the volume, turns off the mute state of the mute noise reduction and echo cancellation unit, and the call can continue.

3. An intelligent PSTN phone voice channel switching noise cancellation method according to claim 1, characterized in that, When performing a second dial during a hands-free call state, initially in the hands-free call state, click the numeric keypad on the central control unit APP call interface to send DTMF numbers. The mute noise reduction and echo cancellation unit controls the external line to be muted, ensuring that the pulse noise generated by the electronic switch switching will not be transmitted to the PSTN external line. Specifically: The media audio channel sets the volume to 0, ensuring that the noise generated by the electronic switch switching cannot be heard on the local handset. The electronic switch in the voice channel switching unit operates to connect the hands-free audio channel to the media audio channel. The media audio channel restores the volume, and the media audio channel can play the sound through the hands-free speaker. The APP dials to emit an analog key tone, which the user can hear on the hands-free speaker. After the PSTN number sending is completed, the media audio channel sets the volume to 0, the electronic switch in the voice channel switching unit operates again, and connects the hands-free audio channel to the PSTN line voice channel connected through the mute noise reduction and echo cancellation unit. The media audio channel restores the volume, turns off the mute of the mute noise reduction and echo cancellation unit, and the call can continue.

4. An intelligent PSTN telephone voice channel switching noise cancellation method according to claim 1, characterized in that, The mute noise reduction and echo cancellation unit is provided with a volume control module for implementing mute processing, an echo cancellation module for implementing echo cancellation processing, and a noise reduction module for implementing voice path noise reduction processing. The noise reduction module adopts a dual-threshold and dual-judgment noise reduction method, which has an enhanced noise reduction effect on the sound frequency of the "click" noise generated during voice channel switching on the line. Specifically: Let a certain line signal be s, and the signal after passing through the noise reduction algorithm calculation be r. s(k) is the voice signal of the kth frame, k represents the subscript of the kth frame, its short-time zero-crossing rate is z(k), that is, the proportion of the number of times the signal passes through the 0 point after k frames, and the short-time power is e(k), that is, the signal power of the audio of the kth frame. Take the frame shift as Ts, and then set the zero-crossing rate threshold as zTol and the power threshold as eTol. Let the enhanced voice after noise suppression be r(k) = s(k), and the background noise suppression criterion is: When e(k) < eTol and z(k) > zTol, then s(k) is determined to be a noise signal, and the denoised signal r(k) is set to 0, thus eliminating the noise signal. When the consecutive non-zero segments in the denoised signal r, i.e., r(k1:k2) ≠ 0 from frame k1 to k2 with a length of k2 - k1 and Ts < 20 ms, then let r(k1:k2) = 0 to eliminate the noise in a short period of time. A part of the "click" noise generated by the switching on the line is included in it and can be eliminated here. After the above double threshold and double discrimination, r(k) will no longer contain non-speech interference and transient interference, i.e., the "click" noise generated during switching on the line. eTol is taken as 0.05 times the maximum value of s, and zTol is taken as 75% of the frame length.

Images