JPH0650888B2 - Voice packet communication system - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a voice packet communication system capable of efficiently transmitting voice from a telephone terminal or the like via a packet communication line network.

[Technical background of the invention and its problems]

Recently, it has been attempted to efficiently communicate voice from a telephone terminal or the like via a packet communication network and talk in real time.

However, the packet communication line network was originally developed for data transmission, and is configured for the purpose of surely transmitting the data even if the transmission delay occurs. Therefore, when voice is transmitted from a telephone terminal for conversation in real time by using this packet communication line network, when trying to faithfully transmit all voice, the large transmission delay causes unnatural conversation.

Therefore, conventionally, attempts have been made to prioritize transmission of voice packets as compared with information packets such as computer data, or to sacrifice the sound quality thereof to ensure real-time performance.

However, even if such measures are taken, it is undeniable that in packet communication, the packet transmission required time (transmission delay time) will vary depending on the traffic state of the transmission path and the operating state of each packet processing unit.

Therefore, a buffer memory is provided in the packet receiving unit to sequentially store the received packets, and the received packets are sequentially read from the buffer memory after a predetermined waiting time. The reproduction waiting control by the buffer memory effectively absorbs the variation in the transmission delay time of each received packet. That is, even if a plurality of packets sequentially transmitted from the transmission unit are randomly received with variations in transmission delay time, they are once stored in the buffer memory and then sequentially reproduced after the waiting time described above. In addition, it is possible to reproduce the packet information that absorbs the variation in the arrival time of each packet.

However, if the waiting time is set to a large value, it is possible to absorb even the received packets that have caused large fluctuations in the transmission delay time, but on the other hand, the time until packet reproduction becomes longer and the telephone voice The response delay becomes a problem when talking by packet communication. In other words, real-time conversation is hindered.

On the other hand, if the waiting time is short, the packets that arrive after the waiting time are discarded, so the amount of information that can be received and played back decreases, and in the case of voice communication, the sound quality deteriorates significantly. I will end up. Therefore, the optimum setting of the waiting time is an important issue.

[Object of the Invention]

The present invention has been made in consideration of such circumstances, and an object of the present invention is to suppress the packet discard rate to a predetermined value or less, and to shorten the reproduction waiting time to ensure real-time audio. It is to provide a voice packet communication method that enables communication.

[Outline of Invention]

The present invention separates a generated voice packet into a voiced packet and an unvoiced packet, and transmits only the separated voiced packet. For example, for a cumulative value indicating the number of generated voice packets and a voiced packet, Header information consisting of a sequence number given in sequence only to the voiced packet is transmitted, and the packet reception unit temporarily stores the received voiced packet in the buffer memory and then waits for the playback waiting time. Then, the voice signal including the silent period is reproduced based on the read voiced packet and the header information added to the voiced packet, and the voice signal is added to the received packet at predetermined time intervals, for example. The packet discard rate is obtained from the header information, and according to this packet discard rate, the playback wait time of the received packet is waited for. The matching time is variably controlled during the silent period, for example.

〔The invention's effect〕

Thus, according to the present invention, only voiced packets are transmitted with the header information added, and the packet receiving unit calculates the packet discard rate from the header information added to the received packet, for example, at predetermined time intervals, and Since the reproduction waiting time of the received packet is variably controlled based on the discard rate, the minimum reproduction waiting time that can suppress the packet discarding rate to a predetermined value or less can be easily set appropriately.

Moreover, as described above, since the unvoiced packet is not transmitted, the silent period can be effectively used to perform the variable control of the reproduction waiting time. Further, it is possible to flexibly deal with the traffic fluctuation of the communication line and set the optimum reproduction waiting time.

Therefore, it is possible to effectively perform voice communication without sacrificing real-time property by reducing the packet waiting rate and ensuring voice quality, and shortening the voice packet playback waiting time.

Example of Invention

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a schematic configuration of a packet transmitting unit 1 and a packet receiving unit 2 of a packet communication terminal to which the embodiment system is applied.

The packet transmitting unit 1 basically digitizes the voice input from the voice input unit 11 via the A / D converter 12,
The encoder 13 converts this into a predetermined code sequence, the packet assembling unit 14 assembles it into a voice packet of a predetermined format, and then the packet transmission control unit 15 transmits the packet to the transmission line 3.

Here, the characteristic configuration of the packet transmission unit 1 to which this method is applied is that the voiced / unvoiced determination unit 16 first inputs the voice, and the voice packet generated by the packet assembling unit 14 is a voiced packet or an unvoiced packet. This is the point where it is determined whether there is. Then, according to the determination result of the determination unit 16,
The packet transmission controller 15 transmits only the voiced packet. That is, unvoiced packets are not transmitted.

Further, the packet assembling unit 14 is given a cumulative value corresponding to the number of generated packets from a cumulative number counter 17, and a sequence number output from the sequence number counter 18 in response to the occurrence of voiced packets. . The cumulative number counter 17 is stepped up each time the packet assembling unit 14 generates a voice packet, and generates a packet cumulative value indicating the number of generated packets. The sequence number counter 18 is incremented when the voice packet generated by the packet assembler 14 is a voiced packet and outputs a sequence number indicating the number of voiced packets generated.

The packet assembler 14 assembles a voiced packet by adding the sequence number N and the cumulative value C to the voice data D as shown in FIG. 2, for example. That is, header information including the sequence number and the accumulated value is added to the voiced packet, and only the voiced packet in such a format is transmitted from the packet transmission control unit 15 to the transmission line
Has been sent to 3.

The packet transmitted from the packet transmission unit 1 to the transmission line 3 is received by the packet reception unit 2 after undergoing various delays according to the traffic state of the transmission line 3.

The packet receiving unit 2 basically stores the received packet received by the transmission control unit 21 in the buffer memory 22 once and then reads the received packet after a predetermined waiting time has elapsed. The voice data in the received packet read from the buffer memory 22 is extracted by the packet disassembling unit 23, decoded by the decoder 24, and then voiced from the D / A converter 25 via the voice output unit 26. It is supposed to be output.

Therefore, the packet discard rate determination unit 27 is
At 23, the header information (sequence number and cumulative value) in the received packet that has been packet-decomposed is input, and the transmission packet discard rate is calculated every predetermined time. This packet discard occurs when there is a large transmission delay such as when the packet to be reproduced next is not received even after the reproduction waiting time of the received packet stored in the buffer memory 22 has elapsed. To do.

The packet discard rate P by the packet discard rate determination unit 27
Is, for example, at some point (at the time of the previous packet discard rate calculation)
Let Nt-1 be the serial number of the received packet obtained in step 3, and Nt be the serial number of the received packet at the time of this packet drop rate calculation, which is the time when a predetermined time has passed since the previous packet drop rate calculation time. When the number of received packets reproduced is S, it is calculated as P = 1-S / (Nt-Nt-1).

According to the packet discard rate P calculated by the determination section 27 in this way, the timing control section 28 reads the received packet from the buffer memory 22 according to whether the discard rate P is smaller than (0.01) or not ( The playback waiting time) is variably controlled in stages.

The variable control of the reproduction waiting time based on the packet discard rate is performed using the silent period that occurs during the period in which the unvoiced packet is originally transmitted because the unvoiced packet is not transmitted.

When the packet discard rate is high, the variable control of the playback queuing time lengthens the queuing time by one rank to reduce packet discard and ensure the quality of transmitted voice. Also, if the packet discard rate is low, the waiting time can be
By shortening it by one rank, real-time processing is ensured within the range where the quality of the communication voice does not deteriorate.

The minimum waiting time as a variable control range is set so as not to fall below the shortest packet delay time, and the maximum waiting time is set as a time that does not make the conversation unnatural. . And within this time range,
It is variably controlled in stages. As a result, the minimum reproduction waiting time that can secure the quality of communication voice by suppressing the packet discard rate to a predetermined value or less is adaptively set, and the real-time property is secured.

FIGS. 3 and 4 show the control of the reproduction waiting time. (T) shows the transmission timing of voiced packets V1, V2, V3, (R) shows the reception timing, and (O) shows the reproduction timing. ing.

That is, the packet receiving section 2 basically operates as shown in FIG. Now, as shown in FIG. 3, it is assumed that the voice packet V1 generated at time t1 is received at time t2 and is reproduced from time t3 when a predetermined waiting time dt has elapsed. In this case, since the next voice packet V2 is received before the reproduction of the voice packet V1 is finished, this voice packet V2 is reproduced following the voice packet V1.

Whether or not the voice packets V1 and V2 are continuous is determined by the cumulative value of the header information.

After that, when the next voice packet V3 is received and it is found from the accumulated value that there is an unvoiced packet that has not been transmitted between the voice packet V2 and the voice packet V2, the difference between the accumulated values from the end of the reproduction of the voice packet V2. The voice packet V3 is reproduced after a packet silence period corresponding to.

In this way, the received packet is reproduced with the same timing relationship as the transmitting side, and the fluctuation of the transmission delay is absorbed.

Then, the transmission of voice packets progresses, and, for example, voice packets V6, V7, and V8 are transmitted in sequence as shown in FIG. 4, and the packet discard rate P is obtained at the time of receiving the voice packet V6 in the packet receiving unit 2. .

The packet discard rate P calculated at this point is 0.01%
If it is determined that the reproduction waiting time dt has a margin, the control for shortening the waiting time dt is performed. However, in this case, even if the queuing time is shortened immediately after the reproduction of the voice packet V6 to control the reproduction timing of the next voice packet V7, the voice packet V7 is continuous with the voice packet V6. ,
The variable control cannot be performed.

Therefore, in the timing control unit 28, a voice packet which is not continuous with the voice packet being reproduced (in this example, the voice packet V8
) Is detected, and the waiting time dt is variably controlled by utilizing the silent period formed during the detection.

As a result, for example, the voice packet V8 is reproduced at time t8 after the waiting time dt shortened from the reception time t7. Thus, the reproduction waiting time dt is variably controlled by using the silent period, and the minimum waiting time dt that can suppress the packet discard rate to a predetermined value or less is set.

Note that this control may not secure a silent period for one packet, but since it is a period in which there is virtually no audio data, it has little effect on the reproduced audio.

As described above, according to this method, the quality of communication voice can be ensured, and the reproduction waiting time of the received packet can be adaptively set to the minimum and reproduced in real time. Therefore,
It enables effective conversation by minimizing the playback delay of communication voice. In addition, it is possible to effectively deal with the fluctuation of the traffic condition of the transmission line, which is a great practical effect.

The present invention is not limited to the above embodiment. For example, each time a predetermined number of packets are received, the packet discard rate may be obtained and the reproduction waiting time may be variably controlled. Further, the variable control of the waiting time may be performed during the voiced packet period by allowing some deterioration in the quality of communication voice. Also, the method of calculating the packet discard rate can be variously modified according to the format of the header information. Further, a threshold value for the packet discard rate for variably controlling the waiting time may be set according to the communication specifications. Further, the transmission line 3 has been described as a packet communication line network, but a communication line using a packet procedure may be applied to an internal network. In addition, voiced / unvoiced determination can be performed even after encoding.
It may be before conversion. In short, the present invention can be variously modified and implemented without departing from the scope of the invention.

[Brief description of drawings]

FIG. 1 is a diagram for explaining an embodiment system of the present invention. FIG. 1 is a schematic configuration diagram of a packet transmitting unit and a packet receiving unit, FIG. 2 is a diagram showing an example of a voice packet format, FIG. 3 and FIG. Each of the figures is a timing chart showing reproduction control of a received packet. 1 …… Packet transmitter, 2 …… Packet receiver, 3 ……
Transmission line, 14 ... Packet assembly unit, 15 ... Packet transmission control unit, 16 ... Voiced / unvoiced determination unit, 17 ... Cumulative number counter, 18 ... Sequence number counter, 22 ... Buffer memory, 23 ... Packet disassembly unit, 27 ... Packet discard rate determination unit, 28 ... Timing control unit.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Katsuhiko Kuma 1 Komukai Toshiba-cho, Kouki-ku, Kawasaki-shi, Kanagawa Inside Toshiba Research Institute Co., Ltd. (56) References JP-B-58-21459 (JP, B1)

Claims (5)

[Claims] 1. In a voice packet communication system for digitizing a voice signal and then transmitting the packetized packet, a packet transmitting unit separates a generated voice packet into a voiced packet and an unvoiced packet, and separates the voiced packet. Header data is added to the voiced packet and transmitted, and the packet reception unit temporarily stores the received voiced packet in the buffer memory and then reads it after the playback waiting time has elapsed. And a voice signal including a silent period based on the header information added to the voiced packet, and the discard rate of the transmission packet is obtained from the header information added to the read voiced packet. According to the above, the voice packet communication system is characterized in that the reproduction waiting time is variably controlled. 2. The voice packet communication system according to claim 1, wherein the variable control of the reproduction waiting time of the received packet is performed during a silent period. 3. The variable control of the reproduction waiting time of a received packet is changed stepwise, and is controlled so as not to exceed a predetermined range. Voice packet communication method. 4. The voice packet communication system according to claim 1, wherein the packet discard rate is calculated based on the number of received packets and the header information thereof at predetermined time intervals. 5. The voice packet according to claim 1, wherein the header information is made up of a cumulative value indicating the number of voice packets generated and a sequence number sequentially assigned only to voiced packets. Communication method.