KR20020037183A - Device and Method for Conference Call Switching - Google Patents

Abstract

PURPOSE: A conference call junction device is provided to reduce delay time for a conference call by one frame by a method of deciding whether PCM voice data is for the conference call, without storing the PCM voice data in a voice memory, so as to supply a good conference call sound quality. CONSTITUTION: An input control block(212) distinguishes data requesting a conference call from data not-requesting the conference call among voice PCM parallel data, and transmits the data in different paths. Input conference voice memories(214) receive the data requesting the conference call, and store the data. A conference control register(215) stores a conference group requesting the conference call, subscribers included in the conference group, channel information and gain information assigned in each subscriber. A gain control and decoding block(216) controls a gain of the stored data, and decodes the data as voice information. A conference synthetic memory(218) synthesizes voice information of the subscribers and stores the synthesized information. An encoding block(219) codes voice information subtracted from the voice information of the subscribers as PCM voice data. Output conference voice memories(220) store PCM voice data modulated in the encoding block(219) by each subscriber requesting the conference call. Voice memories(221) store information stored in the output conference voice memories(220) with PCM voice information not-requesting the conference call.

Description

Device and Method for Conference Call Switching

The present invention relates to an apparatus and method for conferencing a call, and more particularly, to an apparatus and method for minimizing a delay occurring during a conference call by directly processing a conference call block without storing the data requesting the conference. It is about.

Figure 1 shows the configuration of a device for a conference call configured in a conventional commercial chip.

As shown in FIG. 1, the apparatus for a conventional conference call parallels PCM serial data obtained by modulating voice information of a subscriber coming from a pulse code modulation ("PCM") highway. Parallel) an input buffer 101 for converting data, a voice memory 102 for storing parallel data converted by the input buffer, a connection memory 107 for storing switching information of voice information stored in the voice memory, a conference call And a conference control memory 106 that stores information for controlling the conference call block.

The conference call block 112 includes a noise removing block 103 for removing noise for use of information stored in the voice memory 102 in the conference call block, and a decoding block for demodulating PCM data stored in the voice memory into voice information. 104, a data memory 108 for storing the voice information demodulated by the decoding block, a conference synthesis memory 105 for adding voice information of various subscribers for a conference call, and the processed data for the conference call. Encoding block 109 for modulating back to PCM data for storage in speech memory 102, and gain adjustment block 110 for adjusting the voice gain of the PCM data modulated by encoding block 109.

Referring to Fig. 1, the operation of a conventional conference call device will be described.

PCM data for the subscriber's voice information is transmitted over the PCM highway in the form of serial data. The transmitted serial PCM data is input to the input buffer 101, which converts the serial PCM data into parallel PCM data.

Information converted into parallel PCM data through the input buffer 101 is stored in the voice memory 102. The voice memory 102 stores PCM data by providing a separate storage space for each subscriber. For example, in case of supporting 256 subscribers, voice memory 102 has up to 256 addresses, and PCM data is stored in address 1 for subscriber 1 and address 3 for subscriber 3, respectively. The parallel PCM data converted through the input buffer is stored in the voice memory 102 without distinguishing whether it is data for requesting a conference call or data for a general call.

Of the information stored in the voice memory 102, information of a subscriber who has requested a conference call is stored in the conference control memory 106. The meeting control memory stores the meeting group number of subscribers requesting a meeting, the number of subscribers requesting a meeting, and information on the channel assigned to the meeting. For example, if a subscriber set to 1, 23, 120 requests a conference call, the conference group number of the subscribers is 1, subscribers 1, 23, 120 belonging to the conference group, and three channels 1,2,3 assigned to the conference. Stored in the conference control memory 106. Through the control information stored in the conference control memory 106, the conference call block 112 starts data processing for the conference call. This is described as follows.

First, the subscriber who requested the conference is found through the information stored in the conference control memory, and the voice PCM data corresponding to the subscriber is read from the voice memory 102. The read data is the noise removing block in the conference call block 112. First go through (103). The noise removal block 103 serves to remove noise and attenuate the PCM data. After the noise is removed, the decoding block 104 decodes the read PCM data into the actual voice data. The voice information converted through the decoding block is stored in the data memory 108. The data memory 108 stores voice information according to the conference call channel. That is, the subscriber's voice information allocated to channel 1 is stored in data memory 1 and the subscriber's voice information allocated to channel 2 is stored in data memory 2. The voice information of the subscriber is stored in the data memory and the voice information is also stored in the conference synthesis memory. The conference synthesis memory serves to sum voice information of subscribers belonging to a conference group. The method of combining voice information is as follows.

First, subscriber information of channel 1 is stored in the data memory 108 and the conference synthesis memory in the manner described above. If there are a total of three subscribers in a conference group and each subscriber is assigned channels 1, 2, and 3, the same process is performed for channel 2 subscribers after performing the above procedure for channel 1 subscribers. However, while the voice information of channel 2 subscriber is stored in the data memory, the conference synthesis memory combines the voice information of channel 1 subscriber stored in the previous step with the voice information of channel 2 subscriber stored in the current data memory. will be. The same process is performed for channel 3 subscribers, and voice information of three subscribers is added to the conference synthesis memory three times.

After the voice information of all subscribers belonging to one conference group is summed in the conference synthesis memory 105, a process of subtracting the original voice information of one subscriber stored in the data memory is performed from the combined voice information. If the voice information of subscribers 1, 2, and 3 is stored in the conference synthesis memory, the voice information of subscriber 1 is subtracted from the voice data to be sent to subscriber 1. This is necessary because the provision of voice for a conference call is inevitably caused by a time delay. If the voice information is provided together with the time delay, a voice echo phenomenon occurs. Therefore, the voice information to be provided to the first subscriber is information obtained by subtracting the voice data of the first subscriber from the information stored in the conference synthesis memory, and the voice information to be provided to the second subscriber is the information obtained by subtracting the voice of the second subscriber. 102 will be stored for each subscriber. Voice data that is not PCM encoded is stored in the voice memory, and too large voice data may be provided when multiple voices are combined. Therefore, the PCM modulation is performed in the encoding block 109 in the conference block again, and the gain adjusting block 110 is adjusted to adjust the size of the voice data. Adjust the size appropriately.

The voice information stored in the voice memory is once again adjusted through the external gain adjusting block 111, and converted into serial data through the output buffer 112 again.

According to the conventional conference call device as described above, a delay of one frame when storing entered data in the voice memory, a delay of one frame when entering the conference call block from the voice memory, and processing for a conference call in the conference call block. There is a total of five frames, including one frame delay when sending a message, one frame delay when writing conference data back to voice memory, and one frame delay when sending information written to voice memory. However, there is a disadvantage that the conference caller may cause a sense of delay due to delay.

In the present invention, to solve the problems of the prior art as described above, it is proposed a conference call apparatus and method that can minimize the delay of voice information for conference call processing.

Figure 1 shows the configuration of a device for a conference call configured in a conventional commercial chip,

Figure 2 shows the configuration of the conference bonding apparatus according to a preferred embodiment of the present invention,

3 shows a flowchart of a conference call joining method according to a preferred embodiment of the present invention;

4 illustrates the structure of an input control register according to a preferred embodiment of the present invention;

5 illustrates a configuration of a conference control register according to a preferred embodiment of the present invention.

In order to achieve the above object, the conference call joining apparatus according to the present invention comprises: an input control block for distinguishing between the data requesting the conference call and the unsolicited data among the voice PCM parallel data of the subscriber and transmitting them through different paths; An input conference voice memory for receiving and storing the PCM parallel data input to the input control block from the input control block if the PCM parallel data is data requested for a conference call; A conference control register for storing a conference group requesting a conference call, subscribers belonging to the conference group, channel information and gain information allocated to each subscriber; A gain adjustment and decoding block for adjusting a gain of data stored in the input conference speech memory and decoding it into speech information; A conference voice memory for storing voice information requesting a conference call decoded through the gain adjustment and decoding block; A conference synthesizing memory stored in the conference voice memory and storing sum of voice information of subscribers belonging to a conference group; An encoding block for coding voice information obtained by subtracting voice information of a subscriber to provide voice information from the information stored in the conference synthesis memory into PCM voice data; An output conference voice memory for storing the PCM voice data modulated through the encoding block for each subscriber requesting a conference call; And a voice memory for storing each subscriber along with PCM voice information that does not require a conference call to output the information stored in the output conference voice memory.

On the other hand, the conference call joining method according to the present invention comprises the steps of (a) converting the voice PCM serial data input from the PCM highway into voice PCM parallel data; (B) determining whether the converted voice PCM data is data for requesting a conference call; (C) storing the data in the memory for voice output if the data does not require a conference call as a result of the determination in step (b); As a result of the determination in step (b), if the data requesting the conference call is stored in the memory for conference call processing, the conference group information, the subscribers belonging to the group, and the subscribers are allocated to the register for the conference call processing control. Storing the channel information (d); (E) adjusting the gain of the voice PCM data stored in the memory for conference call processing in step (d) and decoding it into voice information; Storing (f) the speech information decoded in step (e); (G) storing the voice information of subscribers belonging to one conference group among the voice information stored in the step (f); Subtracting the voice information corresponding to the subscriber provided from the sum of the voice information in step (g) and encoding the same into PCM data (h); And storing (i) the data encoded in the step (h) and transmitting the stored data to a memory for the voice output of the step (c) in the next frame.

Hereinafter, with reference to the accompanying drawings will be described in detail a preferred embodiment of the conference call joining apparatus and method according to the present invention.

Figure 2 shows the configuration of the conference bonding apparatus according to a preferred embodiment of the present invention.

As shown in FIG. 2, the conference splicing apparatus according to the preferred embodiment of the present invention includes an input buffer 211, an input control block 212, an input control register 213, a conference call block 225, and two voice memories. 221.

The conference call block includes a conference control register 215, two input conference voice memory 214, a gain adjustment and decoding block 216, two conference voice memory 217, a conference synthesis memory 218, an encoding block ( 219) and two output conference voice memories.

In FIG. 2, the input buffer 211 converts serial PCM voice data input from a PCM highway into parallel PCM data as in the conventional apparatus. The input control block 212 determines whether the converted parallel PCM data is the data for which the conference call is requested or not. If the data is the request for the conference call, the input control block 212 transmits the data to the input conference voice memory 214. In case the data does not require a conference call, it transmits the data to the voice memory 221. The input control block has an input demux 226 and an input mux 227, as shown in FIG.

As described above, the input conference voice memory 214 stores data for requesting a conference call among the PCM data input to the input control block 212. The reason why the input conference voice memory is composed of two memories 214a and 214b will be described later. The conference control register 215 also stores a conference group that requested a conference call, subscribers belonging to the conference group, channel information and gain information assigned to each subscriber.

The gain adjustment and decoding block 216 adjusts the gain of the PCM data requesting the conference call stored in the input conference voice memory 214 and decodes it into voice information. The conference voice memory 217 stores voice information decoded by the gain adjustment and decoding block 216.

The conference synthesis memory 218 serves to add voice information to subscribers belonging to the same conference group among voice information stored in the conference voice memory 217. The method of combining voice information in the conference synthesis memory will be described later. When the voice information of one conference group summed in the conference synthesis memory is provided to the subscriber, the voice information corresponding to the subscriber of the summed voice information should be removed. This delay is inevitable due to the nature of the conference call block, since an echo phenomenon occurs for the subscriber's voice unless the voice of the subscriber to which the voice information is provided is removed from the combined voice information.

The encoding block 219 is synthesized in the conference synthesizing memory, and serves to encode voice information, in which speech information corresponding to a subscriber provided therein, is removed, into PCM parallel data. The encoded PCM parallel data is stored in the output conference voice memory 220. The reason why the output conference voice memory is composed of two will be described later.

In the voice memory 221, parallel PCM data that does not require a conference call and parallel PCM data that is processed through a conference call block by requesting a conference call are stored for each subscriber. The gain adjustment block 223 adjusts the gain of the PCM data stored in the voice memory 221. In particular, since the voice data requesting the conference call is a sum of voices of several subscribers, it is necessary to reduce the gain to some extent so that the subscriber does not hear too loudly. The output buffer converts the parallel PCM data into serial PCM data and outputs it to the PCM highway as it was when inputted again.

3 shows a flowchart of a conference call joining method according to a preferred embodiment of the present invention.

Referring to Figure 2 and 3 will be described a conference call joining method according to a preferred embodiment of the present invention.

Serial PCM voice data transmitted from the PCM highway is converted into parallel PCM data through the input buffer 211 (S300). The converted parallel PCM data is stored in the input control register 213, and the input control block 212 determines whether the parallel PCM data is a data requesting a conference call based on the stored information (S301). 4 illustrates the structure of an input control register according to a preferred embodiment of the present invention. As shown in Fig. 4, in the input control register 213, data for which a conference call is requested is displayed with 0 in the set display, and data otherwise is displayed with 1 in the clear display. The input control register 213 also stores a conference channel number, which is used as the address of an input voice memory in which voice information of a subscriber requesting a conference call will be stored later.

The input control block 212 causes the data indicated by '0' to request a conference call to be transmitted to and stored in the input conference voice memory 214 in the conference call block 225 via the input demux 226 (S302). In operation S308, data marked as '1' that does not require a conference call is transmitted to the voice memory 221 through the input mux 227 and stored. As shown in Fig. 2, the input conference voice memory 214 is composed of two memories 214a and 214b, where the PCM data transmitted through the input control block is stored in one of the two memories. In another memory, PCM data requesting a conference call transmitted in a previous frame is stored for processing in the conference block. According to a preferred embodiment of the present invention, the data requesting the conference call is stored in the input conference voice memory 214 at the address according to the conference channel information stored in the input register. For example, if subscribers 1, 20, and 30 requested a conference call in one conference group, if subscribers 1, 20, and 30 were assigned channels 1, 2, and 3, respectively, the voice PCM data of each subscriber was entered. It is stored in address 1, 2, and 3 of conference voice memory.

FIG. 5 illustrates a configuration of a conference control register according to a preferred embodiment of the present invention. As shown in FIG. 5, the conference control register 215 includes a conference group requesting a conference call and a subscriber belonging to the conference group. Channel information and gain information allocated to each subscriber are stored, and the information stored in the conference control register is continuously utilized in processing for conference call.

The information stored in the input conference speech memory is transmitted to a gain adjustment and decoding block 216, and the gain adjustment and decoding block 216 adjusts the gain of the stored PCM data (S303), and converts the PCM data into actual voice information. To decode (S304).

The voice information of the subscriber who requested the decoded conference call is stored in the conference voice memory 217. Voice memories of subscribers belonging to one conference group of stored voice memories are synthesized in conference synthesis memory 218 (S305). Similar to the storage method of the input conference voice memory 214, the voice data decoded by the address according to the assigned channel is stored in the conference voice memory. The conference synthesis memory 218 stores the synthesized voice data according to the address according to the conference group of subscribers who requested the conference call.

A method of synthesizing voice information of subscribers belonging to one conference group in the conference synthesis memory is as follows.

As in the above example, when subscribers 1, 20, and 30 require conference calls, and channels 1, 2, and 3 are assigned to each subscriber, and their meeting group is called 1, subscriber 1 Since the decoded voice information is stored in address 1 of the conference voice memory 217, and subscriber 1 belongs to the group 1, the voice information of subscriber 1 is stored in address 1 of the conference synthesis memory 218. Thereafter, voice information of subscriber 20 is stored in address 2 of the conference voice memory 217, and the stored voice information is synthesized in the conference synthesis memory 218 with the voice information of subscriber 1 stored in the address 1. The synthesized voice information of subscriber 1 and subscriber 20 is stored in address 1 of the conference synthesis memory. The same process as described above is performed with respect to subscriber 30. Consequently, synthesized voice information of subscribers 1, 20, and 30 is stored in address 1 of the conference synthesis memory.

When the synthesized voice information is provided to subscribers 1, 20, and 30 as it is, due to a delay for conference call processing, each subscriber is provided with voice information that causes an echo phenomenon. Therefore, it is preferable that the first subscriber is provided with voice information from which the voice information of the first subscriber is removed from the synthesized voice information, and the second subscriber is provided with voice information from which the voice information of the subscriber 20 is removed. Therefore, after speech synthesis of subscribers belonging to one conference group is completed in the speech synthesis memory 218, a process of removing speech information of each subscriber from the synthesized speech information is performed.

After removing the subscriber's voice information to provide the voice information from the synthesized voice information, the encoding block 219 encodes it back to PCM parallel data (S306). After being encoded with the PCM parallel data, it is stored in the address according to the channel assigned to each subscriber of the output conference voice memory 220 (S307). The reason why the output conference voice memory 220 is composed of two memories 220a and 220b in FIG. 2 is that one of the two memories stores the PCM data encoded in the encoding block, and the other memory encodes the previous frame. This is because, when transferring the data to the voice memory, a memory for processing the data is required.

The PCM data stored in the output conference voice memory 220 is transferred to the voice memory 221 through the input mux 227 of the input control block 212 and stored in the next frame (S308). The reason why the voice memory 221 is composed of two memories 221a and 221b in FIG. 2 is a memory for storing currently transmitted PCM data, and a process of transmitting data transmitted in a previous frame to the gain adjusting block 223. This is because it requires two pieces of memory to process.

The PCM data stored in the voice memory 221 is transmitted to the gain adjusting block 223 to adjust the gain (S309), and the output buffer 224 is converted back into serial data to provide the subscriber with the adjusted PCM data. Convert (S310). The converted data is again output to the PCM highway.

According to the above-described apparatus and method, a delay of one frame when entering the input conference voice memory from the input control block 211, a delay of one frame when processing the conference call, and the processed data are outputted to the voice memory 221. There is a delay of one frame when saving to a frame, and a total of four frames of delay when exporting information stored in the voice memory. Therefore, the delay time for a conference call is reduced.

As described above, according to the conference call joining apparatus and method according to the present invention, a conference call is placed through a method of determining whether or not the conference call is requested data without storing the PCM voice data in the voice memory. Since the delay time can be shortened by one frame, there is an advantage of providing a conference call voice of good quality.

Claims (6)

An input control block for distinguishing between data for requesting a conference call and data not for request among the voice PCM parallel data of a subscriber and transmitting them through different paths; An input conference voice memory for receiving and storing the PCM parallel data input to the input control block from the input control block if the PCM parallel data is data requested for a conference call; A conference control register for storing a conference group requesting a conference call, subscribers belonging to the conference group, channel information and gain information allocated to each subscriber; A gain adjustment and decoding block for adjusting a gain of data stored in the input conference speech memory and decoding it into speech information; A conference voice memory for storing voice information requesting a conference call decoded through the gain adjustment and decoding block; A conference synthesizing memory stored in the conference voice memory and storing sum of voice information of subscribers belonging to a conference group; An encoding block for coding voice information obtained by subtracting voice information of a subscriber to provide voice information from the information stored in the conference synthesis memory into PCM voice data; An output conference voice memory for storing the PCM voice data modulated through the encoding block for each subscriber requesting a conference call; And a voice memory for storing each subscriber with PCM voice information which does not require a conference call to output the information stored in the output conference voice memory. The method of claim 1, The input control block includes an input demux for transmitting voice PCM information requesting a conference call to the input conference voice memory, and an input mux for transmitting voice PCM information not requesting conference call to the voice memory. Conference call splicing device. The method of claim 1, The input conference voice memory includes two memories for storing voice PCM data transmitted from the input control block and a memory for storing voice PCM data transmitted from the input control block in a previous frame for conference call processing. Conference call splicing device, characterized in that. The method of claim 1, The output conference speech memory includes two memories, a memory for storing voice PCM data transmitted from the encoding block, and a memory for storing voice PCM data transmitted from the encoding block in the previous frame to the voice memory. Conference call splicing device, characterized in that. Converting the voice PCM serial data input from the PCM highway into voice PCM parallel data (a); (B) determining whether the converted voice PCM data is data for requesting a conference call; (C) storing the data in the memory for voice output if the data does not require a conference call as a result of the determination in step (b); As a result of the determination in step (b), if the data requesting the conference call is stored in the memory for conference call processing, the conference group information, the subscribers belonging to the group, and the subscribers are allocated to the register for the conference call processing control. Storing the channel information (d); (E) adjusting the gain of the voice PCM data stored in the memory for conference call processing in step (d) and decoding it into voice information; Storing (f) the speech information decoded in step (e); (G) storing the voice information of subscribers belonging to one conference group among the voice information stored in the step (f); Subtracting the voice information corresponding to the subscriber provided from the sum of the voice information in step (g) and encoding the same into PCM data (h); And (i) storing the encoded data in step (h) and transmitting the stored data to a memory for voice output of step (c) in the next frame. . The method of claim 5, And storing data for the conference call in the memory for the conference call of step (d) is stored at a address according to a channel assigned to each subscriber requesting the conference call.