JP2000059391A - Audio packet transmitting/receiving method, audio packet transmission terminal and audio packet reception terminal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the cutoff of voices or quality degradation by reproducing audio packets through interpolation to preceding and following temporary packets to be stored at the time of detecting a lost packet based on transmission audio packet numbers which are sequentially assembled as audio packets, transmitted to an asynchronous digital communication network while added the transmission packet numbers and added to the audio packets. SOLUTION: The two on more temporary packets combined and extracted as prescribed are assembled as one audio packet, successively assembled as audio packets at a packet assembly part 104 and transmitted from a signal transmission part 105 onto an asynchronous digital communication network 2 at high speed later. The audio packets from the asynchronous digital communication network 2 are successively received by a signal reception part 106 and when the number of received audio packets does not reach a prescribed number as a result of packet receiving state monitor from a signal receiving state monitoring part 109, continuity is maintained by reproducing these packets through interpolating processing to preceding and following audio data values to be temporarily stored in an encoded data assembly part 108.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice packet transmitting / receiving method when a voice packet from a voice packet transmitting terminal is received / reproduced by a voice packet receiving terminal via a data communication network, and further, the voice packet transmitting / receiving method. The present invention relates to a voice packet transmitting terminal and a voice packet receiving terminal having a configuration suitable for carrying out the method.

[0002]

2. Description of the Related Art With the development of communication technology and the spread of communication equipment in the field of information communication, asynchronous digital communication networks (A
Transmission and reception of video, audio, and the like are performed via a (TM network), and among them, audio is currently transmitted and received in a packetized state. The voice signal digitally encoded in advance is sequentially divided into voice packets of a fixed length and transmitted from the voice packet transmitting terminal to the asynchronous digital communication network, while the voice packet from the asynchronous digital communication network receives the voice packet. These are sequentially received by the terminal and reproduced as audio.

Here, the voice packet transmitting / receiving method according to the prior art will be specifically described as follows. That is, FIG. 4 shows an outline of a conventional voice packet transmitting / receiving method. In this case, in a voice packet transmitting terminal, an analog voice signal from a voice input unit (microphone or the like) 301 is converted into a digital signal. After being coded and converted into digital voice data by the coding unit 302, the voice data is sequentially assembled as voice packets by the packet assembling unit 303. The voice packets sequentially obtained from the packet assembling unit 303 are transmitted from the signal transmitting unit 304 to the asynchronous digital communication network. On the other hand, in the voice packet receiving terminal, voice packets from the asynchronous digital communication network are sequentially received by the signal receiving unit 305, and the data storage unit 3
06 is temporarily stored. The audio packet temporarily stored in the data storage unit 306 is thereafter read out to the digital audio data decoding unit 307 periodically and in a predetermined order, and then decoded, so that the audio output unit (such as a speaker) ) 308 reproduced and output as audio.

[0004] Incidentally, the voice packet received from the signal receiving section 305 is temporarily stored in the data storage section 306 for the following reason. That is, when voice packets are transmitted / received via the asynchronous digital communication network, not all voice packets are necessarily transmitted via the same path in the asynchronous digital communication network. Depending on the situation, there is a possibility that the data will be transmitted via various detour paths. Therefore, each voice packet may be received by the voice packet receiving terminal with various transmission delays (time delays) according to the path length. Therefore, after each received voice packet is temporarily stored in a FIFO manner, it is periodically read out in a predetermined order.
When decrypted and played back, when playing back as audio,
Reproduction of sound as a state including a discontinuous state (interrupted state) can be prevented.

In the case of Japanese Patent Application Laid-Open No. 9-27827, upon receiving and reproducing an audio packet, the next audio packet is still received due to loss of the audio packet itself during transmission (due to a header error or the like) or a large transmission delay. If not, interpolation is performed on the next voice packet based on the voice packet received immediately before,
The next voice packet is newly reproduced,
The audio can be reproduced without interruption while the transmission delay is reduced and the transmission data amount is reduced.

[0006]

However, in the voice packet transmission / reception method according to the prior art, a certain voice packet is transmitted to the voice packet receiving terminal due to a congestion state on the asynchronous digital communication network or a loss of the packet itself. When the transmission delay until the transmission is large and the voice packet is discarded as a lost packet, interruption of the reproduced voice is inevitable. FIG. 5 (A),
(B) and (C) respectively show transmitted voice data, voice data at the time of packet loss, and reproduced voice data as examples, but no voice data exists in the section corresponding to each lost packet. And as a result, FIG.
As shown in (C), when audio is reproduced and output from the audio output unit 308, the audio is reproduced and output in a state including breaks. Further, in the case of the above publication, even if the sound can be reproduced without interruption, quality deterioration of the reproduced sound is inevitable.

A first object of the present invention is to provide a voice packet receiving terminal which discards a voice packet as a lost packet due to a voice packet loss or a congestion state on an asynchronous digital communication network. Another object of the present invention is to provide a method of transmitting and receiving a voice packet in which the voice can be reproduced and output in a state where the voice is not interrupted and the quality deterioration in reproduction is suppressed. A second object of the present invention is to provide a voice packet transmitting terminal having a configuration suitable for implementing the voice packet transmitting / receiving method. A third object of the present invention is to provide a voice packet receiving terminal having a configuration suitable for implementing the voice packet transmitting / receiving method.

[0008]

The first object of the present invention is to provide a voice packet transmitting terminal in which an analog voice signal is encoded and converted into digital voice data and then divided at regular time intervals. The digital audio data is further provisionally divided into fixed lengths, and the provisionally divided digital audio data is extracted as predetermined packets, and two or more of the provisional packets are extracted and separated in a predetermined interval in a predetermined combination. Each set of two or more temporary packets extracted in combination is sequentially assembled as a voice packet,
While the voice packet number is transmitted to the asynchronous digital communication network in a state where the voice packet number is added, the voice packet receiving terminal generates a temporary packet from the voice packet every time a voice packet is received from the asynchronous digital communication network. Two or more extracted packets are temporarily stored together with other temporary packets already temporarily stored in a state edited in the order of the temporary packet position within a certain time based on the transmission voice packet number added to the voice packet. In parallel with the above, as a result of monitoring the reception state of the voice packet from the asynchronous digital communication network, the number of received voice packets within the predetermined time has not reached the predetermined number, and the order of the transmitted voice packet number is missing. If a certain voice packet is specifically detected as a lost packet, the temporary packet before and after temporarily stored The temporary packets included in the lost packet are reproduced by the interpolation process for the packet, and each of the temporary packets temporarily stored as a state in which continuity in time is maintained is periodically read out sequentially. The above is achieved by reproducing and outputting as audio in a decoded state.

A second object of the present invention is to provide an audio input unit for converting audio into an analog audio signal,
A digital signal encoding unit that encodes and converts an analog audio signal from the audio input unit into digital audio data; and a digital signal encoding unit that divides the digital audio data from the digital signal encoding unit at regular time intervals. The digital voice data is further provisionally divided into fixed lengths, and the provisionally divided digital voice data is used as provisional packets, and two or more provisional packets of the provisional packets are separated by a predetermined distance and extracted in a predetermined combination. And a set of two or more temporary packets extracted and combined in a predetermined manner by the coded data division unit, and assembles them as voice packets, and transmits them to the asynchronous digital communication network in a state where transmission voice packet numbers are added. This is achieved by providing at least a signal transmission unit.

A third object of the present invention is to provide a signal receiving unit which sequentially receives voice packets from an asynchronous digital communication network, and that the signal receiving unit receives the voice packets every time the signal receiving unit receives the voice packets. An encoded data extraction unit that extracts two or more temporary packets from the voice packet based on the transmission voice packet number added to the voice packet, assuming that the position of the temporary packet within a fixed time is specified;
Asynchronous with an encoded data assembling section which temporarily stores two or more temporary packets extracted from the encoded data extracting section in the order of the temporary packet position within a certain time together with other temporary packets already temporarily stored. A signal reception state monitoring unit for constantly monitoring the reception state of voice packets from the digital communication network as the number of voice packets received within the predetermined time period, and a packet reception state monitor from the signal reception state monitoring unit As a result, when the number of received voice packets does not reach the predetermined number and a voice packet due to a missing order of the transmitted voice packet number is specifically detected as a lost packet, the encoded data assembling section temporarily stores A digitizer that reproduces a temporary packet included in the lost packet by performing an interpolation process on the stored temporary packets before and after. An audio data interpolator; a digital audio data decoder for decoding each temporary packet periodically read from the encoded data assembler in a state where continuity in time is maintained; and a digital audio data decoder. This is achieved by providing at least an audio output unit for sequentially reproducing and outputting decoded temporary packets from the decoding unit as audio.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIG.
This will be described with reference to FIG. First, an audio packet transmitting / receiving method according to the present invention will be described. FIG. 1 shows an outline of the audio packet transmitting / receiving method. In this case, in the voice packet transmitting terminal 1, an analog voice signal from a voice input unit (microphone or the like) 101 is coded and converted into digital voice data by a digital signal coding unit 102. At 103, the digital audio data is further divided at regular time intervals t, and the digital audio data within the fixed time period t is further temporarily divided into fixed lengths. As
Two or more of these temporary packets are extracted in a predetermined combination as being separated by a predetermined interval. The two or more temporary packets extracted in a predetermined combination are assembled into one packet by a packet assembling unit 10.
4, the signal transmission unit 1 assembles the voice packets in sequence and adds the transmission voice packet number thereto.
05 is transmitted over the asynchronous digital communication network 2 at a high speed.

As described above, the encoded data dividing section 103
Then, each of the digital audio data provisionally divided within the predetermined time t is a temporary packet, for example, two temporary packets separated from each other by three temporary packets in time so as to be assembled as one voice packet. Each of the temporary packets within t is extracted in a predetermined combination, and voice packets are sequentially assembled in the encoded data division unit 103. Before describing the configuration and operation on the voice receiving terminal side, the temporary packets are extracted. The following is a more specific supplementary description of an example of packet combination extraction.

That is, the total number of digital voice data within a given time t is T (arbitrarily selectable as a value common to voice transmitting / receiving terminals), and the number of voice packet divisions for the digital voice data is n ( Similarly, if the transmission voice packet number is s (arbitrarily selectable on the voice transmission terminal side) and the transmission voice packet number is s (arbitrarily selectable on the voice transmission / reception terminal side), the digital signal is coded by the digital signal coding unit 102. From the digital voice data, a voice data value corresponding to a sampling time obtained as, for example, i * n + s is extracted as a temporary packet. Where i is i = 0,
1,..., (T / n) -1). More specifically, FIG.
(B) shows that the values of T and n are 8, 4, respectively,
Although a specific example in which s = 0, 1, 2, 3 is shown, the number of digital voice data (temporary packets) included in each voice packet is 2 (= T / n) from the values of T and n. , And the possible values of i are obtained as 0 and 1, respectively. Therefore, every time the value of s is sequentially updated from 0 to 3, the value of i is updated to 0 and 1, respectively, and is substituted into i * n + s, so that it is included in the packet for the transmission packet number s. The audio data values can be extracted as a combination as shown.

As can be seen from the above description, in this example, it is assumed that two voice data values separated by four sampling times are extracted in combination, and four voice packets are sequentially generated from digital voice data within a certain time t. Although not limited to this, in general, the coded data dividing section 103 divides the digital audio data from the digital signal coding section 102 at regular time intervals t, and Is temporarily divided into fixed lengths, and the temporarily divided digital voice data is temporarily extracted as predetermined packets, and two or more of the temporary audio packets are separated by a predetermined distance and extracted in a predetermined combination. Good thing.

Now, based on the configuration and operation of the voice packet transmitting terminal 1, the voice packet receiving terminal 3
In the voice packet receiving terminal 3, voice packets from the asynchronous digital communication network 2 are sequentially received by the signal receiving unit 106. Each time a voice packet is received, the encoded data extraction unit 107 specifies a sampling time position within a certain time t from the voice packet based on the transmission voice packet number s added to the voice packet. In this state, two audio data values separated by four sample times are extracted. Further, the encoded data assembling unit 108 edits the two audio data values extracted by the encoded data extraction unit 107 in the order of sampling time positions within a certain time t, and stores the other audio data values already stored temporarily. It is temporarily stored in a FIFO manner together with the data value.

On the other hand, in parallel with the above-mentioned receiving operation, the signal receiving state monitoring unit 109 also checks the receiving state of the voice packet from the asynchronous digital communication network 2 at predetermined time intervals for the voice received within the predetermined time. The number of packets is constantly monitored. At the time of receiving the voice packet, in this example, normally, the transmission voice packet number s is 0 → 1
→ 2 → 3 → 0 → 1 → 2 → 3 → 0...…...
As a result of monitoring the packet reception status from the signal reception status monitoring unit 109, a voice packet in which the number of received voice packets has not reached the predetermined number and which is missing from the order of the transmitted voice packet numbers is specifically detected as a lost packet. In this case, the digital audio data interpolation unit 110
The audio data value included in the lost packet is reproduced, though incomplete, by interpolation processing on the preceding and following audio data values already temporarily stored in the encoded data assembling unit 108. As described above, the audio data value included in the lost packet is reproduced incompletely, and the audio data temporarily stored in the encoded data assembling unit 108 is maintained so that continuity in time is maintained. After being inserted as desired between the data values, the digital audio data decoding unit 11
1, when the data is decoded periodically after being read, the audio can be reproduced and output from the audio output unit (such as a speaker) 112 without interruption irrespective of the presence of a lost packet. is there.

Here, how to reproduce the voice data value included in the lost packet will be described more specifically as follows. For example, assuming that a voice packet corresponding to the transmission voice packet number 1 (including voice data values corresponding to the sample times 2 and 6) shown in FIG.
6, the audio data values corresponding to each of the sampling times 2 need to be reproduced. Among them, the audio data values corresponding to the sampling times 2 are obtained from the audio data values corresponding to the sampling times 1 and 3, respectively.
The audio data value corresponding to the sample time 6 may be reproduced from the audio data values corresponding to the sample times 5 and 7 as, for example, an intermediate value of the audio data values. Also, assuming that a voice packet corresponding to the transmission voice packet number 2 (including voice data values corresponding to the sampling times 3 and 7) is also detected as a lost packet simultaneously with the transmission voice packet number 1 , Sampling time 2,
It is necessary to reproduce the audio data values corresponding to each of 6, 3, and 7, and among them, the audio data values corresponding to each of the sampling times 2 and 3 are obtained from the audio data values corresponding to each of the sampling times 1 and 4. The audio data values corresponding to the sampling times 6 and 7 may be reproduced from the audio data values corresponding to the sampling times 5 and 8 as, for example, proportional distribution values of the audio data values.

FIGS. 3A to 3C show transmission voice data, voice data at the time of packet loss, and reproduced voice data, respectively, as examples in connection with the present invention. As can be seen, each voice packet is discarded as a lost packet, and even if voice data sporadically exists, the lost voice data is reproduced as a state in which deterioration is more suppressed than those voice data. , And can be reproduced and output as audio.

[0019]

As described above, according to the first aspect, a voice packet is discarded as a lost packet at a voice packet receiving terminal due to a voice packet loss or a congestion state on an asynchronous digital communication network. Thus, a voice packet transmission / reception method capable of reproducing and outputting voice without interruption of voice and in a state in which quality deterioration during reproduction is suppressed is obtained. In this case, a voice packet transmitting terminal and a voice packet receiving terminal each having a configuration suitable for implementing the voice packet transmitting / receiving method are obtained.

[Brief description of the drawings]

FIG. 1 is a diagram showing an outline of a voice packet transmitting / receiving method according to the present invention;

FIGS. 2A and 2B are diagrams showing a combination extraction method in one example of provisionally divided digital audio data according to the present invention within a fixed time;

FIGS. 3 (A) to 3 (C) show the case where voice packets having a large reception delay are discarded as lost packets in the voice packet receiving terminal, but are included in the lost packets. Diagram for explaining that sound data can be reproduced and then reproduced as sound

FIG. 4 is a diagram for explaining a voice packet transmission / reception method according to the related art.

FIGS. 5A to 5C are diagrams for explaining a problem in a case where a voice packet having a large reception delay is discarded as a lost packet in a voice packet receiving terminal.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Voice packet transmission terminal, 2 ... Asynchronous digital communication network, 3 ... Voice packet receiving terminal, 101 ... Voice input part,
Reference numeral 102: a digital signal encoding unit; 103, an encoded data dividing unit; 104, a packet assembling unit; 105, a signal transmitting unit; 106, a signal receiving unit; 107, an encoded data extracting unit; 109: signal reception state monitoring unit, 110: digital voice data interpolation unit, 11
1. Digital audio data decoding unit 112: Audio output unit

Claims (3)

[Claims] At a voice packet transmitting terminal, an analog voice signal is coded and converted into digital voice data,
While being divided every fixed time, the digital audio data within the fixed time is further provisionally divided into a fixed length, and the provisionally divided digital audio data is respectively provided as provisional packets.
Two or more of the temporary packets are separated and separated by a predetermined interval in a predetermined combination, and the predetermined combination of the two or more temporary packets are sequentially assembled as voice packets, and the transmission voice packet number is While the voice packet is transmitted on the asynchronous digital communication network as an added state, the voice packet receiving terminal extracts two or more temporary packets from the voice packet each time a voice packet is received from the asynchronous digital communication network. Above, as a state of being edited in the order of the temporary packet position within a certain time based on the transmission voice packet number added to the voice packet, while being temporarily stored together with other temporary packets already stored in parallel, As a result of monitoring the reception status of the voice packet from the asynchronous digital communication network, Wattage is not reached the predetermined number,
When a certain voice packet due to a loss in the order of the transmission voice packet number is specifically detected as a lost packet, the voice packet was included in the lost packet by interpolation processing of the temporary packets before and after temporarily stored. Each temporary packet temporarily stored as a state in which temporal continuity is maintained while the temporary packet is reproduced is periodically read out sequentially, and then reproduced and output as audio in a decoded state. Voice packet transmitting and receiving method. 2. The voice packet transmitting terminal according to claim 1, wherein said voice input unit converts voice into an analog voice signal, and said digital voice code converts said analog voice signal from said voice input unit into digital voice data. A signal encoding unit, and while the digital audio data from the digital signal encoding unit is divided at regular time intervals, the digital audio data within the fixed time is further temporarily divided into fixed lengths. Each of the data is assumed to be a temporary packet, and two or more of the temporary packets are separated by a predetermined distance in a predetermined combination and extracted in a predetermined manner.
An audio packet transmitting terminal comprising: a signal transmitting unit that assembles each of the above-mentioned set of temporary packets as an audio packet and transmits the packet to the asynchronous digital communication network in a state where a transmission audio packet number is added. 3. The voice packet receiving terminal according to claim 1, wherein: a signal receiving section for sequentially receiving voice packets from the asynchronous digital communication network; and each time the voice packet is received by the signal receiving section, An encoded data extraction unit for extracting two or more temporary packets from the voice packet based on the transmission voice packet number added to the voice packet in a state where a temporary packet position within a predetermined time is specified; An encoded data assembling section for temporarily storing the temporary packets extracted from the encoded data extracting section in the order of the temporary packet positions within a predetermined time together with other temporarily stored temporary packets, and asynchronous digital communication A signal reception state monitor that constantly monitors the reception state of voice packets from the network as the number of voice packets received within the predetermined time at predetermined time intervals. As a result of the packet reception status monitoring from the monitoring unit and the signal reception status monitoring unit, a voice packet in which the number of received voice packets has not reached the predetermined number and there is a missing in the order of the transmission voice packet number is identified as a lost packet. A digital audio data interpolator for reproducing the tentative packet included in the lost packet by interpolation processing for the tentative packet before and after temporarily stored in the coded data assembling unit when the erroneous packet is detected; A digital audio data decoding unit for decoding each of the temporary packets periodically read sequentially from the encoded data assembling unit while maintaining the above continuity; and decoding from the digital audio data decoding unit. An audio output unit for sequentially reproducing and outputting the temporary packets as audio.