JP3092037B2 - Voice information multiplex transmission method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice information multiplex transmission method for multiplexing a plurality of voices by fixed length packets and transmitting the multiplexed voice over a communication line.

[0002]

2. Description of the Related Art Recently, multiplex communication using packets has been widely used, but the quality of voice information is a problem.

Hereinafter, a conventional voice information multiplex transmission method will be described with reference to the drawings. FIG. 2 is a schematic diagram showing the concept of a conventional voice information multiplex transmission method. In the figure,
Reference numeral 11 denotes voice information coding means provided for each voice channel, and 12 denotes voice packet multiplexing means for multiplexing voice packet information generated by the voice information coding means of each voice channel and sending the multiplexed packet to a communication line. And
Reference numeral 13 denotes a communication line, 14 denotes voice packet demultiplexing means for distributing voice packet information received from the communication line to each output channel, and 15 denotes voice information of voice packets demultiplexed and provided for each voice channel. Is an audio information decoding means for generating an output audio from the audio information.

[0004] A conventional audio information multiplex transmission method having the above configuration will be described. In the transmitting side device, the audio information encoded by the audio information encoding means 11 has a sound / voice at the time when one packet of information is accumulated for each audio channel.
Silence is determined. When it is determined that the voice information of one packet has sound, a voice packet is configured with one channel of voice information for each channel.
At this time, identification information for identifying the voice channel is added to each packet. The voice packet multiplexing means 12 multiplexes the voice packets generated by each voice information coding means 11 in the order in which they were generated, and sequentially transmits them on the communication line 13.

[0005] Voice packet demultiplexing means 1 of the receiving apparatus
In step 4, the voice packet received from the communication line 13 is multiplexed and separated by the voice channel identification information assigned to each voice packet to the voice information decoding means 15 of each voice channel and distributed. Each audio information decoding unit 15 receiving an audio packet from the audio packet demultiplexing unit 14
Reproduces a source audio signal by extracting and decoding voiced audio information in an audio packet.

As described above, even with the conventional voice information multiplex transmission method, it is possible to encode and packetize only the voiced portion of the voice input of a plurality of channels and multiplex the voice input over a communication line.

[0007]

In such a conventional voice information multiplex transmission method, voice information corresponding to the voice packet length is stored for each voice channel, and then voice / non-voice determination is performed to form a voice packet. Multiplex transmission on a communication line. Therefore, there is a problem that the delay for storing the voice information for one voice packet length is large, and in particular, the voice quality is deteriorated as the coding rate of voice is lower.

An object of the present invention is to solve the above-mentioned problems, and an object of the present invention is to provide a voice multiplex transmission method capable of reducing voice information accumulation delay time per channel and having good voice quality.

[0009]

According to the present invention, in order to achieve the above-mentioned object, a transmitting apparatus multiplexes voice information of a plurality of voice channels into an information part of a fixed-length packet and transmits the multiplexed information to a communication line. In the audio information multiplexing transmission in which individual audio information is extracted from the packet by demultiplexing in the receiving device and audio information of the original audio channel is generated, in the transmitting device, a fixed unit synchronized between the audio channels is used. At each time, a state is shown in which sound / non-speech of voice information input and stored from each voice channel to the respective voice information encoding means within the unit time is determined to indicate voice / non-voice of each voice channel. A state pattern representing information is generated, and a synchronization packet composed of a synchronization pattern representing the unit time interval and the state pattern is transmitted via a voice packet multiplexing unit. The data is transmitted over a communication line, and the audio information within the unit time of the audio channel determined to be sound is arranged in the information portion of the packet in a predetermined audio channel order to generate a fixed-length audio packet. Then, by sequentially sending out the necessary number of the voice packets on the communication line to store all the voiced voice information, the voice information of the voiced voice channel input within one unit time is transmitted. Is completed, and by repeating this operation, the voiced voice information is sequentially transmitted.
Voice packet demultiplexing means receives the synchronization packet, determines the reception timing of the subsequently received voice packet based on the synchronization pattern, detects the arrangement of voice channels in the received voice packet based on the state pattern, and Receives a voice packet,
The voice packets are sequentially demultiplexed in accordance with the order of the detected voice channels and distributed to the voice information decoding means corresponding to the voice channel to output the voice data for one unit of time. This is an audio multiplex transmission method in which multiplexed audio information is sequentially received by repeating this operation after completion.

[0010]

According to the present invention, in the above configuration, the transmitting apparatus transmits a synchronization packet prior to a voice packet, and the voice packet is composed of only voice information of a voiced voice channel and transmitted. Along with the timing, the arrangement order of the voiced voice information in the voice packet is given to the receiving side device. Since voice packets do not contain synchronization information, fixed-length packets can be used only for voice information,
The accumulation delay time during which audio information stays in the audio information encoding means is short. The receiving device knows the reception timing of the voice packet from the synchronization packet, knows the arrangement order of the voice channels in the voice packet, and demultiplexes based on the arrangement.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the voice information multiplex transmission method of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic diagram illustrating the concept of the present embodiment. In the figure, reference numeral 1 denotes voice information coding means provided for each voice channel, and 2 denotes a packet formed by multiplexing voice information generated by voice information coding means of each voice channel. Voice packet multiplexing means 4 for transmitting voice packet information received from a communication line to each output channel;
Is audio information decoding means provided for each audio channel to generate audio information from the audio information of the demultiplexed audio packets. Reference numeral 6 denotes a synchronization packet, and reference numeral 7 denotes a voice packet.

Hereinafter, the interrelationship and operation of the above components will be described. (1) In FIG. 1 (a), the audio information encoding means 1 of each channel determines whether the audio information input within the unit time is sound or silence at every fixed unit time synchronized between the channels. Is determined, and if there is sound, the voice packet multiplexing means 2 is notified that there is voice information to be transmitted. In the present embodiment, the unit time is set to one half of the time length of inputting audio information for one audio packet from one channel. The voice packet multiplexing means 2 scans the above-mentioned notification with respect to the voice information coding means 1 of each channel, and generates a state pattern expressing a voiced / silent state for each channel as a bit string. In a certain unit time shown in FIG. 1, since channels 1, 2, 4, and 6 are sound and have audio information to be transmitted, the state from channel 1 to channel 6 is changed to " Bit patterns “110101” are shown as bit patterns of “voice = 1” and “silence = 0”, and a bit string “110101” in which these are arranged in channel order is generated as a state pattern. As shown in FIG. 1, the voice packet multiplexing means 2 generates a synchronization packet 6 by adding the state pattern following a predetermined constant synchronization pattern indicating a unit time delimiter. Send up.

The voice packet multiplexing / demultiplexing means 4 receives the synchronization packet 6 and knows a unit time delimiter for a voice packet sequence continuously transmitted on the communication line 3 and receives the synchronization pattern "1". By examining the bit position of the audio packet 7, it is determined from which channel the audio information of the audio packet 7 continuing from the synchronization packet 6 comes from.

(2) Next, in FIG. 1 (b), the voice packet multiplexing means 2 selects only the channels having sound and the voice information to be transmitted to the communication line 3 from channels 1 to 6. Audio information is taken out in order to form an audio packet 7. At this time, the unit time is set with respect to the fixed length of the voice packet so that one voice packet 7 is configured from voice information of a plurality of channels. As described above, the unit time in this embodiment is set to one half of the time during which the audio information for one audio packet length is stored. Can store audio information. For this reason, the first audio packet 7 carries the audio information of channel 1 and channel 2, and the second audio packet 7 carries the audio information of channel 4 and channel 6, respectively. , In this order on the communication line 3.

In the voice packet demultiplexing means 4,
Since it is known from the state pattern of the previously received synchronization packet 6 that the voice information of the four sound channels arrives as voice packets 7, the individual voice information is separated from the voice packets 7 which subsequently arrive sequentially. Thus, the audio information can be distributed to the audio information decoding means 5 corresponding to channel 1, channel 2, channel 4, and channel 6.

By repeatedly performing the above operation for each unit time, only the necessary number of audio packets 7 for the sound channel in each unit time are formed and multiplexed and transmitted. The demultiplexing means can decode the signal into a channel corresponding to the sound channel and output audio information.

As described above, according to the audio information multiplex transmission method of the present embodiment, the transmitting side apparatus determines the presence / absence of audio input / output for each fixed unit time synchronized with a plurality of audio channels. The operation of transmitting the state information as a state pattern together with the synchronization pattern, and then generating and transmitting an audio packet 7 composed of only voiced audio information and not including channel identification information is repeated. After detecting the arrangement order of voiced voice channels in the voice packet 7 according to the state pattern, the voice packet 7 is received and demultiplexed, so that the channel information portion in the voice packet is used for transmitting voice information. It can be used effectively and the transmission interval between channels can be shortened so that it can be transmitted one after another. It is possible to reduce the time delay such as to stay in the storage device of the broadcast encoding means, it is possible to better voice information transmission quality.

In the present embodiment, the case where the number of input channels is 6 and the vocal sound information of two channels is multiplexed and transmitted in one voice packet has been described. However, in other cases, the same method is used. It goes without saying that voice information can be multiplexed and transmitted.

[0019]

As is apparent from the above description, according to the present invention, the transmitting apparatus multiplexes voice information of a plurality of voice channels into the information section of a fixed-length packet, transmits the multiplexed voice information to the communication line, and In the audio information multiplexing transmission in which individual audio information is extracted from the packet by demultiplexing in the device and audio information of the original audio channel is generated, the transmitting device transmits the audio information every fixed unit time synchronized between the audio channels. In the unit time, voice / non-speech of the voice information input and stored from each voice channel to the respective voice information coding means is determined, and state information indicating voice / non-voice of each voice channel is represented. A state pattern is generated, and a synchronization packet 6 composed of a synchronization pattern indicating the unit time interval and the state pattern is transmitted to a communication line via a voice packet multiplexing unit. Then, the voice information within the unit time of the voice channel determined to have sound is arranged in the information portion of the packet in a predetermined voice channel order to generate a fixed-length voice packet, The number of the voice packets necessary to store the voiced voice information is sequentially transmitted over the communication line, so that the voice information of the voiced voice channel input within one unit time can be transmitted. When the operation is completed, this operation is repeated so that the voiced voice information is sequentially transmitted. In the receiving apparatus, the voice packet multiplexing / demultiplexing means receives the synchronization packet 6 and subsequently receives the synchronization packet 6 according to the synchronization pattern. The reception timing of the voice packet is determined, and the arrangement of the voice channels in the received voice packet is detected based on the state pattern. , And sequentially demultiplexes the audio packets in accordance with the order of the detected audio channels, and distributes and outputs the audio packets to audio information decoding means corresponding to the audio channels. By completing the reception of the information and repeating this operation, the multiplexed audio information is sequentially received, so that the delay time required for storing the audio information and configuring the audio packet can be reduced, and the sound quality can be reduced. Good voice information multiplex transmission is possible.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a configuration of an embodiment of a voice information multiplex transmission method of the present invention.

FIG. 2 is a schematic diagram showing a configuration of a conventional voice information multiplex transmission method.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 voice information encoding means 2 voice packet multiplexing means 3 communication line 4 voice packet demultiplexing means 5 voice information decoding means 6 synchronization packet 7 voice packet

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) H04J 3/00-3/26 H04L 12/28 H04L 12/56

Claims (1)

(57) [Claims] 1. A transmitting device multiplexes voice information of a plurality of voice channels into an information portion of a fixed-length packet and transmits the multiplexed information to a communication line. The receiving device multiplexes and separates each voice information from the packet. In the audio information multiplexing transmission for extracting and generating audio information of the original audio channel, the transmitting side apparatus transmits each audio signal from each audio channel within the unit time at every fixed unit time synchronized between the audio channels. It determines the presence / absence of sound in the audio information input and stored in the information encoding unit, generates a state pattern representing state information indicating the presence / absence of sound / non-speech of each audio channel, and indicates the unit time segment. A synchronization packet composed of a synchronization pattern and the state pattern is sent out onto a communication line via a voice packet multiplexing unit, and then a voice determined to be voiced is output. The audio information in the unit time of the channel is arranged in the information portion of the packet in a predetermined audio channel order to generate a fixed-length audio packet, and the number of the audio information necessary to store all the voiced audio information is generated. By sequentially transmitting the voice packets onto the communication line, the transmission of the voice information of the voice channel input in one unit time is completed, and the voice voice information is repeated by repeating this operation. The transmission is performed sequentially, and the receiving device
Voice packet demultiplexing means receives the synchronization packet, determines the reception timing of the subsequently received voice packet based on the synchronization pattern, detects the arrangement of voice channels in the received voice packet based on the state pattern, and Receives a voice packet,
The voice packets are sequentially demultiplexed in accordance with the order of the detected voice channels and distributed to the voice information decoding means corresponding to the voice channel to output the voice data for one unit of time. A voice multiplex transmission method in which multiplexed voice information is sequentially received by repeating the above operation.