JP2012124689A - Communication system, transmission side device, reception side device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce sound distortion or sound skipping when an allowable storage amount of jitter absorption buffer increased at the time of communication network congestion is returned to a default value after the communication network congestion is recovered.SOLUTION: A transmission side device 1 comprises: a silent state information generating unit 11 which generates silent state information in accordance with a volume level of voice data; a voice packet generating unit 12 which adds the silent state information to packetize the voice data; a packet delivering unit 13 which transmits a packet to a reception side device 3. The reception side device 3 comprises: a packet receiving unit 31 which receives the packet to acquire the voice data and the silent state information; a jitter absorption buffer 33 which stores the voice data to adjust a delay time; and a jitter buffer adjusting unit 32 which, when a storage amount exceeds a specified allowable storage amount, increases the allowable storage amount and, when the storage amount is within a specified allowable storage amount for a certain time, discards the voice data indicating a silent state and returns the allowable storage amount back to a default value.

Description

  The present invention relates to a communication system such as a VoIP communication system equipped with a VoIP (Voice Over IP) terminal adapter function for providing an IP (Internet Protocol) telephone service, a transmission side apparatus, and a reception side apparatus.

In recent years, a telephone service provided by a telecommunications carrier uses VoIP technology in which analog voice data is digitized and then converted into IP packets and transferred through the same packet communication network (such as the Internet) as data communication such as a PC. Yes.
When the VoIP technology is used, there is an advantage that an inexpensive service can be provided because a dedicated voice device is not required in the communication network and can be shared with the data communication network. On the other hand, when congestion occurs in the communication network, jitter occurs in the arrival time of the packet at the receiving side device. In this case, when continuous audio data is reproduced by the reception CODEC of the reception side device, there is a problem that sound skipping occurs and the audio quality deteriorates.

  Therefore, in the conventional receiving side apparatus, in order to absorb jitter due to congestion of the communication network, a jitter absorbing buffer is provided in the preceding stage of the receiving CODEC. The received audio data is accumulated up to an allowable accumulation amount in the jitter absorption buffer, and then transferred to the reception CODEC. Thereby, the occurrence of underrun due to jitter can be suppressed.

  However, when jitter further increases, underrun occurs in the jitter absorption buffer, and sound skip occurs. Therefore, when the occurrence of underrun is detected, the allowable accumulation amount in the jitter absorption buffer is increased, and the subsequent underrun occurrence is suppressed to ensure the voice quality.

  Although the voice quality is ensured by the above control, since the delay is added by the increase in the allowable accumulation amount in the jitter absorption buffer, the conversation becomes unnatural. Therefore, when the accumulated amount of the jitter absorbing buffer is maintained within the allowable accumulated amount for a certain time, it is considered that the congestion of the communication network has been recovered, and the allowable accumulated amount of the jitter absorbing buffer is returned to the default value. At this time, the accumulation amount of the jitter absorption buffer is reduced by increasing the reproduction speed in the reception CODEC or unconditionally discarding the transfer data to the jitter absorption buffer (see, for example, Patent Document 1).

Japanese Patent Application Laid-Open No. 2004-282692

  As described above, in the method disclosed in Patent Document 1, when the allowable accumulation amount of the jitter absorption buffer is returned to the default value, the reproduction speed in the reception CODEC is increased, or the transfer data to the jitter absorption buffer is unconditionally set. Is discarded. However, there is a problem that the quality of the sound is affected, such as distortion and skipping of the reproduced sound.

  The present invention has been made to solve the above-described problems. The allowable accumulation amount of the jitter absorption buffer increased when the congestion of the communication network occurs, and the delay optimization is performed after the congestion of the communication network is recovered. Therefore, an object of the present invention is to provide a communication system, a transmission side apparatus, and a reception side apparatus that can suppress influences on voice quality such as voice distortion and sound skipping when returning to default values.

  A communication system according to the present invention includes: a transmission side device that packetizes and transmits audio data; and a reception side device that receives packets from the transmission side device and converts the packets into audio data. A silence state information generating unit that generates silence state information based on the volume level, a voice packet generating unit that packetizes the corresponding voice data by adding the silence state information generated by the silence state information generation unit; A packet sending unit for transferring the packet generated by the voice packet generation unit to the corresponding receiving device, and the receiving device receives the packet transferred by the packet sending unit and acquires voice data and silence state information. A packet receiving unit, a jitter absorbing buffer that accumulates audio data acquired by the packet receiving unit and adjusts a delay time, and a jitter When the amount of audio data stored in the collection buffer exceeds the set allowable storage amount, the allowable storage amount is increased, and the amount of audio data stored in the jitter absorption buffer is set for a certain period of time. And a jitter buffer adjustment unit that discards voice data whose silence state information acquired by the packet reception unit indicates a silence state and returns the permissible accumulation amount to a default value.

  According to the present invention, since it is configured as described above, the allowable accumulation amount of the jitter absorption buffer that is increased when the communication network is congested is set to the default value for delay optimization after the congestion of the communication network is recovered. When returning to, effects on voice quality such as voice distortion and skipping can be suppressed.

It is a figure which shows the structure of the VoIP communication system which concerns on Embodiment 1 of this invention. It is a flowchart which shows the jitter buffer adjustment operation | movement in the VoIP communication system which concerns on Embodiment 1 of this invention. It is a figure which shows reduction of the accumulation amount of the jitter absorption buffer in Embodiment 1 of this invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
Embodiment 1 FIG.
1 is a diagram showing a configuration of a VoIP communication system according to Embodiment 1 of the present invention.
As shown in FIG. 1, the VoIP communication system includes a transmission side device 1, a communication network (such as the Internet) 2, and a reception side device 3.

  The transmission side device 1 digitizes the analog voice data received from the telephone 4 and converts it into an IP packet. The transmission side device 1 includes a transmission CODEC (silent state information generation unit) 11, an audio packet generation unit 12, and a packet transmission unit 13.

The transmission CODEC 11 is an ITU-T G. 711, G. In accordance with a method defined by 729a or the like, analog audio data from the telephone 4 is encoded into digital audio data.
Further, when encoding the analog audio data, the transmission CODEC 11 detects the volume level of the audio data, and compares it with a predetermined threshold value to generate silence state information. Here, if the transmission CODEC 11 determines that the volume level is below the threshold, the transmission CODEC 11 determines that the audio data is in a silent state. The silent state information generated by the transmission CODEC 11 indicating whether the voice data is silent is transferred to the voice packet generator 12 together with the digital voice data.

The voice packet generator 12 converts the digital voice data from the transmission CODEC 11 into IP packets according to RTP (Real-time Transport Protocol) at regular intervals.
Further, when the voice packet generation unit 12 converts the digital voice data into an IP packet, the silence state information transferred together with the digital voice data is converted into a DSCP (DS Code Point) value in a DS (Differentiated Services) area of the IP packet. Convert and embed. Specifically, when the silence state information indicates a silence state, the low-priority packet (for example, DSCP value = 0x22) is embedded, and otherwise, the high-priority packet (for example, DSCP value = 0x2e) is embedded.
The IP packet generated by the voice packet generator 12 is transferred to the packet transmitter 13.

The packet sending unit 13 transfers the IP packet from the voice packet generation unit 12 to the corresponding receiving side device 3 via the communication network 2.
The communication network 2 transfers the IP packet from the transmission side device 1 to the corresponding reception side device 3, and is the same packet communication network as data communication such as a PC.

  The receiving side device 3 receives the IP packet from the transmitting side device 1 via the communication network 2 to acquire digital voice data, converts it into analog voice data, and transmits it to the telephone 5. The reception side device 3 includes a packet reception unit 31, a jitter buffer adjustment unit 32, a jitter absorption buffer 33, and a reception CODEC 34.

  The packet receiver 31 receives an IP packet from the transmission side device 1 via the communication network 2 and acquires digital audio data. In addition, the packet receiving unit 31 examines the DSCP value of the received IP packet and acquires silence state information. The digital audio data and silence state information acquired by the packet receiving unit 31 are transferred to the jitter buffer adjusting unit 32.

  The jitter buffer adjustment unit 32 adjusts the amount of audio data (allowable accumulation amount) accumulated in the jitter absorption buffer 33. Note that the default value of the allowable accumulation amount is set to a value that can absorb the average jitter of the communication network 2, for example. Here, when the jitter buffer adjustment unit 32 determines that the jitter has increased due to the occurrence of congestion in the communication network 2 and has exceeded the amount of jitter that can be absorbed by the set allowable storage amount, the jitter buffer adjustment unit 32 increases the allowable storage amount. . In addition, when the accumulation amount of the jitter absorption buffer 33 is maintained within the set allowable storage amount for a certain time, the jitter buffer adjustment unit 32 returns the allowable storage amount to the default value. At this time, the jitter buffer adjustment unit 32 reduces the accumulation amount of the jitter absorption buffer 33 by discarding the audio data in which the silence state information acquired by the packet reception unit 31 indicates the silence state. Restore the default value.

  The jitter absorption buffer 33 accumulates digital audio data from the packet reception unit 31 via the jitter buffer adjustment unit 32, adjusts the delay time, and transfers it to the reception CODEC 34 to absorb jitter. . The jitter absorption buffer 33 accumulates the digital audio data up to the allowable accumulation amount set by the jitter buffer adjustment unit 32, and then starts the transfer to the reception CODEC 34.

  The reception CODEC 34 decodes the digital audio data from the jitter absorption buffer 33 into analog audio data. The analog audio data decoded by the reception CODEC 34 is transmitted to the telephone 5.

Next, the jitter buffer adjustment operation in the VoIP communication system configured as described above will be described.
FIG. 2 is a flowchart showing the jitter buffer adjustment operation in the VoIP communication system according to Embodiment 1 of the present invention, and FIG. 3 is a diagram showing the reduction of the accumulated amount of the jitter absorption buffer 33 in Embodiment 1 of the present invention. is there.

  Note that the IP packet transmitted from the transmission side device 1 includes silence state information in addition to the digital voice data. That is, when encoding the analog audio data, the transmission CODEC 11 generates silence state information based on the volume level of the audio data. The voice packet generation unit 12 adds silence state information when digital voice data is converted into IP packets.

  In the jitter buffer adjustment operation in the VoIP communication system, as shown in FIG. 2, first, the jitter buffer adjustment unit 32 determines whether an underrun has occurred in the jitter absorption buffer 33 in the normal state (steps ST21 and ST22). . That is, the jitter buffer adjustment unit 32 causes the jitter absorption buffer 33 to increase the jitter due to congestion in the communication network 2 and when the IP packet reception interval continues for a certain period of time at a rate less than the reproduction speed of the reception CODEC 34. It is determined that the amount of jitter that can be absorbed is exceeded, and it is determined that an underrun has occurred.

In this step ST22, when the jitter buffer adjustment unit 32 determines that an underrun has occurred in the jitter absorption buffer 33, the receiving side apparatus 3 shifts to a communication network congestion state and the allowable accumulation amount of the jitter absorption buffer 33 is reached. Is increased (steps ST23 and ST24). In other words, the jitter buffer adjustment unit 32 performs the ITU-T G. By using a PLC (Packet Loss Concealment) function defined by 711 Appendix 1 or the like, the last reproduced audio data is interpolated (in the case of continuous, the second and subsequent amplitudes are reduced by 20%) to prevent sound skipping. Then, the allowable accumulation amount of the jitter absorption buffer 33 is increased (for example, an increase corresponding to the data amount interpolated by the PLC function). In a communication network congestion state, voice data is accumulated in the jitter absorption buffer 33 up to the increased allowable accumulation amount by interpolating the voice data every several packets by the PLC function. After that, when the audio data is accumulated up to the allowable accumulation amount, the interpolation of the audio data by the PLC function is stopped.
By this processing, underrun in the jitter absorption buffer 33 can be prevented.

Next, the jitter buffer adjustment unit 32 detects the accumulation amount of the jitter absorption buffer 33 and determines whether or not it is maintained within the allowable accumulation amount for a certain time (step ST25).
In step ST25, when the jitter buffer adjusting unit 32 determines that the accumulated amount of the jitter absorption buffer 33 exceeds the allowable accumulated amount, the jitter buffer adjusting unit 32 returns to step ST24 and increases the allowable accumulated amount.

On the other hand, when the jitter buffer adjusting unit 32 determines in step ST25 that the accumulated amount of the jitter absorbing buffer 33 is maintained within the allowable accumulated amount for a certain time, the accumulated amount of the jitter absorbing buffer 33 is the default value. Is determined (step ST26). In other words, if the state where the allowable storage amount is increased from the default value continues, skipping will be eliminated by adding the storage amount, but the state where the delay is increased will continue, so the conversation will be unnatural due to deviation. Become. Therefore, the allowable accumulation amount of the jitter absorption buffer 33 is returned to the default value.
In step ST26, when the jitter buffer adjustment unit 32 determines that the accumulated amount of the jitter absorption buffer 33 is the default value, the process returns to step ST21, and the receiving side apparatus 3 shifts to the normal state.

  On the other hand, in step ST26, when the jitter buffer adjustment unit 32 determines that the accumulation amount of the jitter absorption buffer 33 is not the default value, that is, the accumulation amount is larger than the default value, the reception-side apparatus 3 determines that the allowable accumulation amount is sufficient. In order to return the amount to the default value, the process shifts to a delay optimization process state (step ST27).

  When returning the allowable accumulation amount to the default value in the delay optimization processing state, the conventional reception side device increases the reproduction speed in the reception CODEC, or unconditionally discards the transfer data to the jitter absorption buffer. The amount of accumulation in the jitter absorption buffer 33 was reduced, and in both cases, sound distortion and skipping occurred.

  Therefore, the receiving-side apparatus 3 according to Embodiment 1 discards silence data when returning the allowable accumulation amount to the default value in the delay optimization processing state (step ST28). That is, the packet receiving unit 31 examines the DSCP value of the received IP packet and acquires silence state information. Next, the jitter buffer adjustment unit 32 discards silence data based on the acquired silence state information. Here, as shown in FIG. 3A, the jitter buffer adjustment unit 32 transfers the audio data whose silence state information does not indicate the silence state to the jitter absorption buffer 33 as shown in FIG. On the other hand, as shown in FIG. 3B, the audio data whose silence state information indicates the silence state is discarded without being transferred to the jitter absorption buffer 33. As a result, as shown in FIG. 3C, the amount of accumulation in the jitter absorption buffer 33 can be reduced while minimizing the influence on the voice quality.

Next, the jitter buffer adjustment unit 32 determines whether or not the accumulation amount of the jitter absorption buffer 33 is a default value (step ST29).
In this step ST29, when the jitter buffer adjusting unit 32 determines that the accumulated amount of the jitter absorption buffer 33 is the default value, the jitter buffer adjusting unit 32 returns to step ST21, and the receiving side apparatus 3 shifts to the normal state.

  On the other hand, when the jitter buffer adjustment unit 32 determines in step ST29 that the accumulation amount of the jitter absorption buffer 33 is not the default value, that is, the accumulation amount is greater than or equal to the default value, the process returns to step ST28 and the silence data Continue to dispose of. Thereafter, the silent data in the jitter absorption buffer 33 is discarded every several packets, and after the accumulation amount of the jitter absorption buffer 33 returns to the default value of the allowable accumulation amount, the reception side apparatus 3 shifts to the normal state.

  As described above, according to the first embodiment, the transmission-side apparatus 1 detects the silence state based on the volume level of the voice data, adds the silence state information to the IP packet, and transfers it to the reception apparatus. Further, the receiving side device 3 acquires the added silence state information, and when the communication network 2 recovers from the congestion state, only the voice data in the silence state is discarded among the transfer data to the jitter absorption buffer 33. . With this configuration, when the allowable accumulation amount of the jitter absorption buffer 33 is returned to the default value for delay optimization after congestion recovery, it is possible to suppress audio distortion and skipping, and to improve audio quality. The influence can be suppressed to the minimum, and END-END delay optimization can be achieved.

Embodiment 2. FIG.
The voice packet generation unit 12 according to the first embodiment is configured to convert the silence state information into a DSCP value in the DS region of the IP packet and embed it. When the silence state information indicates a silence state, the IP packet is padded. The packet length may require a region, and the silence state information may be embedded in the padding region.
As described above, when the silent state information is added to the IP packet, the use of the padding area can be expected to have an effect that there is no restriction in the service using the priority control based on the DSCP value.

Embodiment 3 FIG.
The voice packet generation unit 12 in the first embodiment is configured to convert the silence state information into a DSCP value in the DS region of the IP packet and embed it. However, when the silence state information indicates a silence state, A (Virtual Local Area Network) tag may be attached, and silence state information may be converted into a COS value of the VLAN tag and transferred as a non-priority packet.
As described above, when the silent state information is added to the IP packet, the same effect as in the second embodiment can be expected by using the COS value of the VLAN.

  In the present invention, within the scope of the invention, any combination of the embodiments, or any modification of any component in each embodiment, or omission of any component in each embodiment is possible. .

  DESCRIPTION OF SYMBOLS 1 Transmission side apparatus, 2 Communication network, 3 Reception side apparatus, 4,5 Telephone, 11 Transmission CODEC (silent state information generation part), 12 Voice packet generation part, 13 Packet transmission part, 31 Packet reception part, 32 Jitter buffer adjustment Part, 33 jitter absorption buffer, 34 reception CODEC.

Claims (9)

In a communication system comprising: a transmission side device that packetizes and transmits voice data; and a reception side device that receives packets from the transmission side device and converts them into voice data;
The transmitting device is:
A silence state information generating unit for generating silence state information based on a volume level of the audio data;
A voice packet generator for packetizing the corresponding voice data by adding the silence state information generated by the silence state information generator;
A packet sending unit for transferring the packet generated by the voice packet generating unit to a corresponding receiving device;
The receiving side device
A packet receiving unit that receives the packet transferred by the packet sending unit and obtains voice data and silence state information;
A jitter absorption buffer for accumulating audio data acquired by the packet receiver and adjusting a delay time;
When the amount of audio data stored in the jitter absorption buffer exceeds the set allowable storage amount, the allowable storage amount is increased, and the amount of audio data stored in the jitter absorption buffer is set for a certain period of time. A jitter buffer adjusting unit that discards voice data indicating that the silent state information acquired by the packet receiving unit indicates a silent state when it is within the allowable storage amount, and returns the allowable storage amount to a default value; A communication system characterized by the above. The communication system according to claim 1, wherein the voice packet generation unit converts and embeds silence state information into a DSCP (DS Code Point) value in a DS (Differentiated Services) area of a packet. The communication system according to claim 1, wherein the voice packet generation unit embeds silence state information in a padding area of a packet. The communication system according to claim 1, wherein the voice packet generation unit adds a VLAN (Virtual Local Area Network) tag to the packet, converts silence state information into a COS value of the VLAN tag, and embeds the silent state information. In the transmission side device that packetizes and transmits audio data,
A silence state information generating unit for generating silence state information based on a volume level of the audio data;
A voice packet generator for packetizing the corresponding voice data by adding the silence state information generated by the silence state information generator;
A transmission-side apparatus, comprising: a packet transmission unit that transfers a packet generated by the voice packet generation unit to a corresponding reception-side device. 6. The transmission side apparatus according to claim 5, wherein the voice packet generation unit converts the silence state information into a DSCP value in a DS region of the packet and embeds it. 6. The transmission side apparatus according to claim 5, wherein the voice packet generation unit embeds silence state information in a padding area of a packet. 6. The transmission side device according to claim 5, wherein the voice packet generation unit attaches a VLAN tag to the packet, converts silence state information into a COS value of the VLAN tag, and embeds it. In the receiving device that receives the packet and converts it into voice data,
A packet receiving unit that receives a packet to which silence state information indicating whether the voice data is silent when the voice data is packetized, and acquires the voice data and the silence state information;
A jitter absorption buffer for accumulating audio data acquired by the packet receiver and adjusting a delay time;
When the amount of audio data stored in the jitter absorption buffer exceeds the set allowable storage amount, the allowable storage amount is increased, and the amount of audio data stored in the jitter absorption buffer is set for a certain period of time. A jitter buffer adjusting unit that discards voice data indicating that the silent state information acquired by the packet receiving unit indicates a silent state when it is within the allowable storage amount, and returns the allowable storage amount to a default value; A receiving apparatus characterized by the above.

Images