JPH0678062A - Voice communication equipment for inter-multipoint - Google Patents

Abstract

PURPOSE:To attain highly efficient communication by adders of high quality or of the small number of address without restricting the frequency bands and coding speeds of signal applied from respective communication terminals to the same contents in respect of voice communication for connecting plural voice communication terminals and executing N-1 voice adding processing. CONSTITUTION:This voice communication equipment is provided with a judging circuit 5 for detecting the frequency band and coding speed of transmitting sound signals applied from respective terminals, decoders 11a to 13a corresponding to the frequency bands and coding speeds of respective terminals, their band converters when necessary, coders 21a to 23a corresponding to the decoders 11a to 13a and the band converters of respective terminals, their band converters when necessary, respective adding means 71, 72 for adding voice-decoded signals respectively applied from N-1 terminals excluding the terminal input of a certain point, finding out the added values of the N points and outputting N-1 added signals as the receiving signal of the point concerned and corresponding to the frequency bands and coding speeds, and selector controlling means 6 selecting one from among decoders 11a to 13a, band converters, and coders 21a to 23a based on a prescribed rule by the output of the circuit 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice conference communication in a case where voice communication between multipoints located at a remote place is such that the coding / decoding characteristics of terminals communicating with each other are not unified. It is a thing.

[0002]

2. Description of the Related Art FIG. 11 shows an example of the configuration of a conference trunk that realizes a conventional multipoint voice communication system. This figure is described in JP-A-63-191453, and the number of conference subscribers is three. The signals from some terminals consist of 8-bit code words and are processed with PCM signals with a sampling frequency of 8 KHz. Depending on other terminals, a sampling frequency of 16 KHz and an ADPCM (Adaptive) having a maximum word length of 4 bits may be used.
Differential PCM-Adaptive Difference PCM
-) Only the signal shall be processed.

In FIG. 11A, code converters (transcoders) C11 and C1 are provided for one first conference subscriber.
2 and decoding converters (trans decoders) D11 and D12
Are arranged. The same applies to the first and third conference subscribers and transcoders C21, C22, C31, C3.
2 and the trans-decoders D21, D22, D31, D3
2 is also arranged. By the way, the transcoders C11, C21 and C31 convert the non-linearly coded 8-bit PCM word into a 14-bit linear code word. At the same time, the sampling frequency is increased to 16 KHz by interpolation and oversampling. The transcoders C12, C22, C32 also convert ADPCM words into linear code words of 14-bit length. The sampling frequency is also 16 KHz in this case. 16 KHz by the trans decoders D11, D21, D31
A 14-bit long linear codeword with a sampling frequency of 4 is subjected to filtering, undersampling, and new quantization to form a non-linearly encoded 8-bit PCM word with a sampling frequency of 8 KHz. Similarly, the trans-decoders D12, D22, D32 cancel the action of the coders C12, C22, C32,
An ADPCM word is formed from a linear code word having a bit length.
The transcoder and the transdecoder are correspondingly called with the following meanings. That is, when the transdecoder in question is connected to the transcoder afterward, the transcoders are called correspondingly in the sense that they coincide with the input signal of the transcoder.

One of the transcoders C11 or C12 is connected to the input side of the adder circuit AD via the changeover switch UE1. Similarly, each transcoder C2
Each one of 1, C22 to C31, C32 is connected to the changeover switches UE2, UE3. Trans decoder D1
1, the input side of D12 is connected to the output side of the adder circuit AD 90, and the trans decoders D21, D22 and D are connected.
The input sides of 31 and D32 are similarly connected. The output side of one of the two transformer decoders D11 and D12 is
It is connected to the output terminal A1 via another changeover switch UA1. Then, the output signal for the first subscriber is transmitted through the changeover switch. The same applies to the trans decoders D21, D22 to D31, D32, the changeover switches UA2 to UA3, and the terminals A2 to A3, respectively.

When the conference connection is established, the signals transmitted from the conference subscribers are applied to the terminals E1 to E3. For example,
The first subscriber sends the information as an ADPCM word having the above format. By a control circuit not shown in FIG. 8, the transcoder C12 is connected to the adder circuit AD90 and the transdecoder D12 is connected to the terminal A1 via the changeover switch UE1. Similarly, for example, an AD having the same format as that described above for the information.
The transcoder C22 and the transdecoder D are also provided for the second subscriber who is supposed to be transmitting and receiving the PCM signal.
22 is selected. The third subscriber is, for example, 8 KHz
Suppose we are processing a non-linearly coded PCM signal consisting of 8-bit codewords of sampling frequency. Third
For the subscriber, the control circuit thus connects the transcoder C31 with the adder circuit AD90 and the transdecoder D31 with the terminal A3. Adder circuit A
D 90 forms a partial sum of its input signals. That is, the output signal of the subscriber circuit AD, intended for one subscriber, is formed from the sum of the input signals of two other subscribers. In the case of this example, the two 14-bit linear codes of the second and third subscribers, which are added to the adder circuit for directing to the first subscriber, are added, and the sum is transmitted to the output side for the first subscriber. To be done.

According to the configuration of FIG. 8B, the transcoders C11 to C32 and the transdecoders D11 to D3 are included.
8 has the same function as that of FIG. 8A, and the intermediate transcoder SC is provided between the first adder circuit 91 and the second adder circuit 92.
193 and SC2 94 are provided. Thereby, each subscriber receives the sum signal of all the conference subscribers matching the respective coding scheme.

A conventional multipoint voice communication system has a configuration as shown in FIG. 8, and the input signal to the multipoint conferencing apparatus is a non-linear PCM having an 8 KHz sampling frequency.
Transcoders C11, C2, both words and ADPCM words with a 16 KHz sampling frequency
1, C31 and C12, C22, C32 all cause the output signal to be a 14-bit long linear code word with a sampling frequency of 16 KHz. Therefore, the input terminals E1, E2. Even if the audio signals input from E3 are all signals having a sampling frequency of 8 KHz, the conventional method expands them into 14-bit linear code words, and then performs 16-bit interpolation and over-sampling processes.
It will be converted into a sampling frequency of KHz.
At least unless each subscriber has the same encoding
It becomes essential to use multiple adder circuits.

[0008]

The conventional multipoint voice communication system corresponds to the decoder, the encoder, the selection switch, and the coding system corresponding to the coding system of the subscriber terminal to be connected. There has been a problem that an adder circuit is required, and in addition, these adder circuits require the simultaneous use of adders of all communicating encoding systems when communicating between terminals of different encoding systems.

The present invention has been made to solve the above problems, and a terminal device connected to a voice communication conference device between multiple points is encoded by a frequency band and an encoding speed of the terminal device. It is an object of the present invention to obtain a flexible and efficient voice communication device by selecting the adding means depending on the case without being fixed to the adding means limited to the decoder and the decoder.

[0010]

A multipoint voice communication apparatus according to the present invention comprises a determination circuit for detecting the frequency band and coding rate of a voice signal transmitted from each terminal, and the frequency band and code for each terminal. A decoder corresponding to the coding speed, a band converter if necessary, an encoder corresponding to the decoder and the band converter for each terminal, a band converter if necessary, and a terminal input at the point. Except for the above, the speech decoded signals from the respective terminals at N-1 points are added, this is obtained for the N points, and each N-1 added signal is output as a received signal at the point. Each corresponding adding means and selector control means for selecting a decoder, a band converter, and an encoder for each terminal based on a predetermined rule by the output of the determination circuit are provided. According to the invention of claim 2, in addition to the invention of claim 1, level adjusting means for decoding the transmission voice signal from each terminal, converting the signal to a predetermined amplitude level and outputting it to the adding means is added. The invention of claim 3 is
In the invention of claim 1, the telephone band voice addition circuit is used as the addition means of the low frequency band.

[0011]

In the multipoint-to-point voice communication apparatus according to the present invention, the frequency band and the coding rate of the voice signal from each terminal are monitored, the coding form is selected from this result and a predetermined rule, and the corresponding adding means is provided. To be elected. Each terminal communicates with the addition means, the decoder and the encoder. According to the invention of claim 2,
Since the voice signal has a level fluctuation at each transmission / reception point, the level becomes relatively the same level after the decoding process and is input to the adding means. In the invention of claim 3, the low-frequency signal adding means is not particularly provided in the multipoint voice communication device, the conventional voice adding circuit of the telephone band is shared, and the low-frequency voice signal is input to the telephone band voice adding circuit. It

[0012]

EXAMPLES Example 1. FIG. 1 is a diagram showing an embodiment of a voice processing unit including a voice encoding / decoding and voice adding unit of a multipoint voice communication apparatus using the present invention. In the figure,
It is assumed that there are two types of input voices, wideband voice and narrowband voice. In the figure, 5 is a determination circuit, 6 is a selector control unit, 11-13 is a decoding circuit, 21-23 is an encoding circuit,
31 to 33 are input side switches, 41 to 43 are output side switches, 71 is a wide band voice addition circuit, 72 is a narrow band voice addition circuit, 81 is a narrow band voice communication path, 82 is a wide band voice communication path, and 83 is a narrow band. A voice communication path, 84 is a broadband voice communication path. The input voice signals VI1 to VIn from the terminal are connected to the switches SWI131 to SWIn33 of the respective channels, and the voice type codes S1 to Sn are input to the determination circuit 5. The decoding circuit 11 is a wideband speech decoder Da.
1 and a narrow band speech decoder Db1 and a band converter Dd1. The output of Da1 is a wide band speech communication path 82.
Connected to. Further, the output from the narrow band voice decoder Db1 through the band converter Dd1 is also the wide band voice communication path 8 in the same manner.
Connected to 2. The output of the narrow band decoder Db1 is connected to the narrow band voice communication path 81. The wideband voice adder circuit 71 is provided between the wideband voice call path 82 and the wideband voice call path 84, and the narrowband voice adder circuit 72 is provided between the narrowband voice call path 81 and the narrowband voice call path 83. . Output of wideband voice adder circuit 71 and narrowband voice adder circuit 7
The output of 2 is input to the encoding circuit 21, and the encoding circuit 21
Is output to the output side switch SWO1. This output is selected by the output side switch SWO1 and becomes an output signal as VO1.

Similarly, an input side switch SWI2 32 and a decoding circuit 12 are prepared for the input signal VI2, and the output side also has an encoding circuit 22 and an output side switch SWO2.
42 is prepared. Below, an input side switch SWIn33 and a decoding circuit 13 are prepared for the input audio signal VIn, and an encoding circuit 23 and an output side switch S are provided on the output side.
WOn 43 is prepared. The switch SWI1 31 and the switch SWO1 41 form a pair, and either the wide band side or the narrow band side is selected by the output of the selector control unit. Each decoding circuit comprises a Da * (* = 1 to n) wideband decoder, a Db * narrowband decoder, and a Dd * upsampling band converter. The encoding circuit paired with this is an Ea * wideband encoder and an Ea * wideband encoder.
It is composed of a narrow band encoder for b * and a band converter Ec * by downsampling.

Next, the operation will be described. In the figure, it is assumed that the input voice signals VI1 and VIn are wideband voice signals and VI2 is a narrowband voice signal. In addition, the voice addition is performed according to the wider band. Input voice signal VI1
In response to this, the selector control unit 6 selects the A contact of the switch SWI1 31 by the output of the determination circuit 5 based on the voice type code. Similarly, a switch S on the output side that makes a pair
WO141 also selects the A side. Therefore, the input voice signal VI1 passes through the decoder Da1 and the wideband voice addition circuit 7
Input to 1. On the other hand, since the input voice signal VI2 has a narrow band, the selector control unit 6 selects the switch SWI2 32 and the switch SWO242 on the B side from the determination circuit output according to the voice type code. In this case, the input audio signal VI2 is input to the wideband audio adding circuit 71 via the decoder Db2 and the band converter Dd2.

Similarly, each of the input signals VIn is input to the wide band voice addition circuit 71 by selecting one of the decoders according to the input voice type. The wideband speech adding circuit 71 adds the signal inputs at the remaining N-1 points excluding the input at each point and returns the result to the coding circuit of the corresponding terminal. This is repeated for each point. The encoding circuit 21 corresponding to the input signal VI1 encodes the output of the wideband speech addition circuit 71 through the wideband encoder Ea1 and outputs it as VO1. Encoding circuit 2 corresponding to input signal VI2
In 2, the output of the wide band speech addition circuit 71 is once down-sampled via the band converter Ec2, coded by the narrow band coder Eb2, and output as VO2.
Similarly, the encoding circuit corresponding to each input signal operates the corresponding encoder to obtain the corresponding audio signal VOn. As described above, in this embodiment, the wide band of the input signal is selected to be operated, and therefore only the wide band voice adding circuit is used.

Example 2. FIG. 2 is a block diagram showing another embodiment. 2, the decoding circuits 11a to 13a are provided with band converters on both the wide band side and the narrow band side, as shown in detail in FIG. Similarly, the encoding circuit 2
1a to 23a are provided with band converters on both the wide band side and the narrow band side, as shown in detail in FIG. Other components are the same as those in FIG. By configuring each decoding circuit and each encoding circuit as shown in FIGS. 3 and 4, whether the input signal is a wideband signal or a narrowband signal,
The wide band or narrow band voice adding circuit can be freely used. The operation of the circuit of FIG. 2 is similar to that of FIG. FIG. 3 is a configuration diagram showing details of the decoding circuit. In the figure, Dax111, D
The functions of bx112, Dcx113 and Ddx114,
The relationship between the input and output signals is shown in FIG. As well
FIG. 5 (b) shows the relationship of the signals processed by. As described above, according to the decoding circuit of FIG. 3, a wide band linear signal and a narrow band linear signal can be arbitrarily generated with respect to both the narrow band input signal and the wide band input signal. FIG. 4 is a configuration diagram showing details of the encoding circuit of FIG. In the figure, E
ax11, Ebx212, Ecx213 and Edx21
4 shows the functions and their input / output signals in FIG.
Shows the conversion form of the signal.

Embodiment 3. Another embodiment will be described. In the configuration of FIG. 2, the selector control unit corresponds to the voice type code by a wideband decoder and a wideband encoder for the wideband input signal, and a selector control unit for the narrowband voice input signal. Is an example in which a narrow band decoder and a narrow band encoder are selected and a band converter is not used. At this time, the switch selection status of the selector control unit 6 in FIG.
FIG. 7 shows a path through which the n speech signals pass through the decoding / encoding circuit.
It shows in (a). In this case, since the band converter is not used, there is no extra distortion or delay, and a good quality audio signal can be obtained. However, both the wideband voice adding circuit 71 and the narrowband voice adding circuit 72 are used as the voice adding circuit.

Example 4. As yet another embodiment, FIG.
In the above configuration, an example will be described in which the output of the selector control unit 6 selects the band having the larger number of input audio signals of the participation conference. In this case, for example, when the number of subscribers of the narrow band voice signal is large, the voice quality is deteriorated with respect to the input of the wide band voice signal, but only the narrow band voice adder circuit 72 may be used as the voice addition circuit. On the contrary, when there are many wideband audio signals as the input audio signals, the selector control unit 6 selects the wideband audio signal side. In this case, no improvement in quality can be expected for the narrow band voice input signal, but only the wide band voice adder circuit 71 is used as the adder circuit. In either case, the use status of the decoding circuit and the encoding circuit is shown in FIG.
It shows in (b).

Example 5. If the input audio signals are all wideband audio signals or all narrowband audio signals,
Alternatively, voice processing can be realized by any of the circuits shown in FIG. In this case, the band converter is not used, and the voice adding circuit may use only one of the wide band voice adding circuit 71 and the narrow band voice adding circuit 72 corresponding to the input and output voice signals. In this case, the switch selection of FIG. 2 and
The operation path of the audio signal in FIGS. 3 and 4 is set as shown in FIG.

Example 6. Further, in the configuration shown in FIG. 2, a system in which the conference organizer selects as the selection criterion of the output of the selector control unit 6 can be adopted. FIG. 8B shows details of switch selection in this case and setting of an operation path of a signal passing through the decoding circuit and the encoding circuit. When the circuit of FIG. 2 is used, in any of the above embodiments, the decoding circuit and the encoding circuit may have the same components for each audio signal, and the parts can be commonly used.

Example 7. FIG. 9 is a block diagram showing an embodiment of the invention of claim 2. In the figure, 91 and 92 are level adjusting circuits. The other components are the same as those in FIG. Next, the operation will be described. The input voice signal from each terminal has a different amplitude level due to the usage state of the equipment at each terminal, transmission loss, and the like. These input signals are once linearized by the decoding circuit, and then the level adjusting circuit 91 for wide band speech and the level adjusting circuit 92 for narrow band speech.
Is adjusted to a constant level and is input to the wideband voice adding circuit 71 or the narrowband voice adding circuit 72, respectively. The subsequent operation is the same as that described with reference to FIG. By doing so, the input voice signal from each terminal becomes the same level, and the influence of fluctuations at each point is suppressed.

Example 8. FIG. 10 is a block diagram for explaining an embodiment of the invention of claim 3. FIG. 10 shows a linear code form of wideband speech as CCIPTREC. G. According to the addition method of multipoint communication recommended by G.722, the circuit configuration is such that N-1 addition is performed separately by converting into linear code for wide band and low band, and in practice, audio band Since the voice addition circuit and the telephone band voice addition circuit are not used at the same time, the low band audio band voice addition circuit can be shared by the telephone band voice addition circuit 73. As a result, the device can be simplified and the circuit can be downsized.

[0023]

According to the multipoint voice communication apparatus of the present invention, a determination circuit for determining the frequency band and coding rate of a voice signal transmitted from each terminal, a decoder and a band converter for each terminal, and An encoder, a band converter, and each addition means corresponding to a frequency band and a coding speed are provided, and a decoder, a band converter, and an encoder are selected based on a determination circuit and a predetermined rule. Therefore, there is an effect that high-quality voice communication can be carried out as necessary, and a highly efficient voice communication device can be constructed by using a small number of voice adding circuits. Further, according to the invention of claim 2, there is an effect that the level difference caused by the environment of each sending end can be eliminated and a voice communication apparatus having no level difference can be realized. According to the invention of claim 3, since the voice signal addition circuit of the telephone band is shared as the low-frequency addition circuit, there is an effect that the device can be downsized.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a multipoint voice communication apparatus according to an embodiment of the present invention.

FIG. 2 is another configuration diagram of the multipoint-to-point voice communication device which is another embodiment of the present invention.

FIG. 3 is a configuration diagram of a decoding circuit in FIG.

FIG. 4 is a configuration diagram of the encoding circuit in FIG.

5 is a diagram illustrating processing and input / output of each element of the decoding circuit in FIG.

6 is a diagram showing processing and input / output of each element of the encoding circuit in FIG.

FIG. 7 is a diagram showing switch settings and signal paths in the third and fourth embodiments.

FIG. 8 is a diagram showing switch settings and signal paths in the fifth and sixth embodiments.

FIG. 9 is a configuration diagram including a circuit for performing level adjustment between respective points, which is an example of the invention of claim 2;

FIG. 10 is a configuration diagram showing an embodiment of the invention of claim 3;

FIG. 11 is a configuration diagram showing an embodiment of a conventional multipoint voice communication device.

[Explanation of symbols]

 5 Judgment circuit 6 Selector control part 11-13, 11a-13a Decoding circuit 21-23, 21a-23a Encoding circuit 31-33 Voice communication device input stage switch 41-43 Voice communication device output stage switch 71 Broadband voice adder circuit 72 Narrowband speech adder circuit 111 Wideband speech decoder 112 Narrowband speech decoder 113 Downsampling band converter 114 Upsampling band converter 211 Wideband speech encoder 212 Narrowband speech encoder 213 Downsampling band converter 214up Bandwidth converter by sampling

Claims (3)

[Claims] 1. A determination circuit for detecting a frequency band and a coding rate of a voice signal transmitted from each terminal, a decoder corresponding to the frequency band and the coding rate for each terminal, and a band converter as necessary. , An encoder corresponding to the above-mentioned decoder and band converter for each terminal, a band converter as necessary, and a speech decoded signal from each terminal of N-1 points excluding the terminal input at the point is added. , This is obtained for N points, and each N-1 addition signal is output as a reception signal at that point. Each addition means corresponding to the above-mentioned frequency band and coding rate, and the output of the above-mentioned judgment circuit make a predetermined rule. A multipoint voice communication device comprising a decoder for each terminal, a band converter, and selector control means for selecting an encoder based on the above. 2. The multipoint speech according to claim 1, further comprising level adjusting means for decoding the voice signals transmitted from the respective terminals, converting the signals to a predetermined amplitude level and outputting the amplitude levels to the adding means. Communication device. 3. The multipoint-to-point voice communication device according to claim 1, wherein a telephone band voice addition circuit is used as the addition means for the low frequency band.