JP2005039724A - Communication control method and communication controller - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication control method and a communication controller capable of maintaining appropriate communication between terminals under communication, suitable for a line state irrespective of variance of bandwidth allocated to the terminals. <P>SOLUTION: A terminal A is equipped with one or a plurality of voice coding systems and performs communication with a voice coding system converter T with SIP control function via a speech codec A. Similarly, a terminal B is also equipped with one or a plurality of voice coding systems and performs communication with the voice coding system converter T with SIP control function via a speech codec B. The voice coding system converter T with SIP control function is equipped with a function for preventing termination of communication by shortage of bandwidth by dynamically switching, mutually converting and relaying the voice coding systems of the terminals A, B in the case of SIP control VoIP communication between the terminals A, B in which the bandwidth to be allocated in the communication changes. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a communication control method and a communication control apparatus for controlling IP voice signal communication using a session start (initialization) protocol between terminals having one or a plurality of voice coding schemes.

  Conventionally, as a communication network in which this kind of communication is performed, there are a mobile communication network and a unified communication network in which different types of communication networks are connected and operated in an integrated manner. Even in such a unified communication network, wireless communication has been promoted in recent years. For example, a plurality of different radio access networks (RAN: Radio Access Network) such as a wireless LAN and a PHS are connected to a common core network (Common Core network). A wireless integrated network that enables seamless communication between these different wireless access networks by connecting to (Network) has also been proposed (see Non-Patent Document 1, for example).

  Here, such a wireless integrated network is usually configured to include a signaling network, a software defined radio (SDR) terminal, a multi-mode wireless terminal, and the like in addition to the common core network. The multi-mode wireless terminal is a wireless terminal that includes a plurality of interfaces and a plurality of applications corresponding to the interfaces, and uses a unique ID (number or URI, etc.) for designating a partner terminal for communication. . Then, whether it is the software radio terminal or the multi-mode radio terminal, when the communication is started, position information and the like are transmitted to the common core network. Accordingly, the common core network side selects a radio access network based on the information, and notifies the software radio terminal or multimode radio terminal of the selected radio access network information. Thus, after the optimum radio access network is selected, information necessary for establishing a session is notified from the common core network to the software radio terminal or the multimode radio terminal.

As described above, in the wireless integrated network, a plurality of wireless access networks are tightly integrated by performing wireless access network selection, session establishment, user information management, and the like through the common core network. A unified communication network is realized.
G. Wu, et al, “MIRAI architecture for heterogeneous networks”, IEEE Com. Mag. , Feb. 2002

  By the way, in such a unified communication network, particularly IP voice signal communication using a session start protocol between terminals having one or a plurality of voice coding schemes, a voice coding scheme common to the terminals is provided. Unless you can communicate directly.

  On the other hand, if both terminals have a common voice coding scheme and the bandwidth necessary for the voice coding scheme used in the communication can be secured in both terminals in the communication establishment phase, Can start. However, after that, if the bandwidth changes and the bandwidth required for the speech coding scheme cannot be secured in either or both terminals, the communication is terminated.

  Note that in IP voice signal communication using the above session start protocol, for example, VoIP communication by SIP control, there is a procedure (RFC3311 or the like) for switching a voice coding method commonly provided between terminals during the communication. . However, as the bandwidth changes, even if the voice coding method is switched between terminals automatically or manually based on these procedures, if either terminal cannot secure the required bandwidth, The communication ends.

  That is, in the above-described conventional communication network, the bandwidths of the least used bandwidths among the voice coding schemes provided in the terminals A and B are Ca and Cb, respectively, and the bandwidths allocated to the terminals are Ba, When Bb is set, communication can be performed only when the following conditional expression (1) is satisfied.

max (Ca, Cb) ≦ min (Ba, Bb) (1)

And once these conditional expressions are not satisfied and communication is interrupted, connection operation via the common core network is required again, and interruption in situations where it is difficult to secure bandwidth in the first place ( In the case of communication end), the connection operation itself becomes difficult.

  Note that the restriction by the conditional expression (1) similarly exists not only in the wireless integrated network but also in a wired unified communication network, and the actual situation based on the restriction is almost common.

  The present invention has been made in view of such circumstances, and the object of the present invention is to make sure that the appropriateness between the terminals in accordance with the line status is appropriate regardless of changes in the bandwidth allocated to the communicating terminals. It is an object of the present invention to provide a communication control method and a communication control apparatus that can maintain simple communication.

  To achieve the above object, the invention described in claim 1 is a communication control method for controlling IP voice signal communication using a session initiation protocol between terminals having one or a plurality of voice coding schemes. While measuring the bandwidth allocated to the terminals in the terminal, among the terminals, a terminal having a plurality of speech coding schemes is dynamically switched according to the measured bandwidth. This is the gist.

  According to a second aspect of the present invention, in the communication control method according to the first aspect, the measurement of the bandwidth is performed by transmitting measurement packets to the communication path of the communicating terminal at regular time intervals. This is done by measuring the round-trip delay and obtaining the bandwidth that can be used as the bandwidth from the sending speed of the measurement packets when a predetermined correlation is detected in the round-trip delay between two consecutive measurement packets. The gist of this is

  According to a third aspect of the present invention, in the communication control method according to the second aspect, the measurement packet size at the start of the bandwidth measurement is a voice notified in advance from a terminal having the plurality of voice encoding methods. Each time a large delay exceeding a predetermined value with respect to the round-trip delay of the measurement packet is detected, the packet size is smaller than that of the speech encoding method of the largest packet size among the encoding methods. The gist of this is to sequentially switch to the same size as that of the speech encoding method.

  According to a fourth aspect of the present invention, in the communication control method according to any one of the first to third aspects, the speech coding method in use between the terminals is required based on the measurement of the bandwidth. A mode in which the dynamic switching of the speech coding scheme is switched to a speech coding scheme with a lower bit rate when a situation in which at least one terminal lacks available bandwidth for a certain bandwidth is detected The gist of what is done in

  According to a fifth aspect of the present invention, in the communication control method according to the fourth aspect of the present invention, when there is no voice coding scheme to be switched in a terminal that requires switching of the voice coding scheme, the corresponding terminal The gist is to temporarily hold the communication between the two.

  The invention according to claim 6 is a communication control apparatus for controlling IP voice signal communication using a session start protocol between terminals having one or a plurality of voice coding schemes, from one of the communicating terminals. First and second speech encoding scheme converters having a function of converting speech encoding data to be transmitted into speech encoding schemes that can be used with the bandwidth allocated to the other terminal, A bandwidth measuring unit that measures the available bandwidth allocated to the terminal, an IP communication processing unit that performs IP communication processing between various terminals, and the communication based on a bandwidth measurement result by the bandwidth measuring unit A speech coding scheme that can be used in the terminal in the terminal is determined, and the corresponding speech coding scheme conversion in the first and second speech coding scheme conversion units is converted to the determined speech coding scheme. A voice coding method control unit that instructs the terminal, a terminal control function based on the session start protocol, and a proxy server function, from the voice coding method control unit to the corresponding voice coding method conversion unit The gist of the present invention is to include a SIP control unit that exchanges a message corresponding to the instructed content with the corresponding terminal via the IP communication processing unit.

  According to a seventh aspect of the present invention, in the communication control device according to the sixth aspect, the bandwidth measuring unit is configured to keep a measurement packet constant in a communication path of the communicating terminal via the IP communication processor. It can be used as the bandwidth from the transmission speed of the measurement packets when a predetermined correlation is detected in the round-trip delay of two consecutive measurement packets, while transmitting at time intervals and measuring the round-trip delay The gist of this is to obtain a sufficient bandwidth.

  The invention according to claim 8 is the communication control apparatus according to claim 7, wherein the bandwidth measuring unit uses the plurality of speech encoding methods as the size of the measurement packet at the start of the bandwidth measurement. The packet size is the same as that of the voice packet coding method with the largest packet size among the voice coding methods notified in advance from the terminal provided, and a large delay exceeding a predetermined value is detected with respect to the round trip delay of the measurement packet. The gist is to sequentially switch to a packet size of the same size as that of the voice encoding method having a packet size smaller than that size each time.

  The invention according to claim 9 is the communication control device according to any one of claims 6 to 8, wherein the speech coding scheme control unit is based on the measurement of the bandwidth by the bandwidth measurement unit, When a situation is detected in which at least one terminal has insufficient available bandwidth with respect to the bandwidth required for the speech coding scheme being used between the terminals, switching the speech coding scheme, The gist of this is to switch to a lower bit rate speech coding scheme.

  According to a tenth aspect of the present invention, in the communication control apparatus according to the ninth aspect, the speech coding scheme control unit is configured to switch the speech coding scheme to be switched to a terminal that needs to switch the speech coding scheme. When there is no system, the gist is to send a message to the terminals via the SIP control unit and the IP communication processing unit in order to temporarily hold communication between the corresponding terminals.

(Function)
According to the first or sixth aspect of the present invention, in IP voice signal communication using a session start protocol between terminals having one or a plurality of voice coding schemes, bandwidths allocated to the terminals in communication The width is measured. And among these terminals, the terminals having a plurality of speech coding schemes are dynamically switched in accordance with the measured bandwidth. For this reason, even if the bandwidth allocated to the terminal in communication changes, the termination of communication due to such a change in bandwidth can be prevented, and it can be matched to the line conditions in each case. An appropriate speech encoding method can be selected.

  According to the second or seventh aspect of the invention, the bandwidth is measured based on the measurement of the round trip delay of the measurement packet transmitted to the communication path at regular time intervals. In addition, at this time, the bandwidth that can be used as the bandwidth is obtained from the transmission speed of the measurement packets when a predetermined correlation is detected in the round-trip delay between two consecutive measurement packets. However, the bandwidth measurement with high accuracy according to the line condition is performed each time.

  According to the third or eighth aspect of the invention, when measuring the bandwidth, the measurement packet size at the start of the bandwidth is a voice coding scheme notified in advance from a terminal having the plurality of voice coding schemes. Each time a large delay exceeding a predetermined value is detected with respect to the round-trip delay of the measurement packet, the speech code having a packet size smaller than that of the speech encoding method having the largest packet size. It is possible to switch to the same size as that of the system. For this reason, the above-described bandwidth measurement is performed more efficiently, that is, in a shorter time, and as a result, the dynamic switching of the speech coding scheme is performed more quickly and appropriately.

  Note that if the measurement packet is transmitted in addition to the voice encoded packet during voice communication, the voice delay causes a shortage of bandwidth due to the measurement packet when the bandwidth is marginal, resulting in trouble with voice communication. May cause. In such a case, the transmission of the measurement packet is interrupted, and the jitter (the difference between the arrival time of the packet and the time stamp embedded in the packet (in the RTP header)) or the packet discard rate is measured. By doing so, the operation of the bandwidth measuring unit can be switched so as to monitor the line state. If the transmission frequency of measurement packets is properly managed, jitter, delay, etc. that adversely affect voice quality can be minimized, but once the original communication path is very thin, it occurs once. This is an operation for relieving such a situation since it may take a long time to eliminate the jitter and the like, and the voice quality may be affected during that time.

  In addition, since encoded speech packets are transmitted at regular intervals (for example, every 20 ms), if the bandwidth is insufficient, packets accumulate at the bottleneck, delaying the arrival time of packets, and increasing jitter. To go. If the jitter becomes so large that it cannot be solved by a jitter buffer or the like, the sound is finally reproduced with a large delay at which the sound is interrupted. Therefore, it can also be determined that a shortage of bandwidth occurs when the jitter exceeds the allowable value.

  Packets accumulated in the bottleneck may be discarded as processing such as congestion control for routers, etc., but since each packet is assigned a sequence number, the discarded packets can be detected. . That is, even when the packet discard rate that hinders voice communication is reached, it can be used as a trigger for switching the voice coding system because the bandwidth is insufficient.

  Furthermore, a measurement method that does not impose a bandwidth by combining active measurement of transmitting a main measurement packet and passive measurement of monitoring the reception state of an auxiliary voice packet is possible. In this case, when a voice packet is not transmitted, it is naturally necessary to transmit a measurement packet.

  According to the invention described in claim 4 or 9, based on the measurement of the bandwidth, at least one bandwidth that can be used with respect to the bandwidth required for the speech coding scheme being used between the terminals is provided. When a shortage is detected in the terminal, the voice coding schemes are dynamically switched in a manner of switching to a lower bit rate voice coding scheme. As a result, the termination of communication due to the aforementioned insufficient bandwidth is suitably suppressed.

  According to the invention described in claim 5 or 10, when there is no speech coding method to be switched to the terminal that needs to be switched, the communication between the corresponding terminals is temporarily suspended. . As a result, it is possible to wait for the resumption of communication based on the bandwidth recovery without performing a troublesome operation such as the reconnection operation described above.

  As described above in detail, according to the present invention, even if the bandwidth allocated to the terminal in communication changes, the termination of communication due to the change in bandwidth is prevented in advance. Thus, IP voice signal communication based on selection of a more appropriate voice coding method suitable for the respective line conditions becomes possible.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, an embodiment embodying a communication control method and a communication control device according to the invention will be described with reference to the drawings.
FIG. 1 is an explanatory diagram for explaining a schematic configuration of a unified communication network 11 capable of VoIP communication by SIP control as a target in the present embodiment. In this unified communication network 11, the terminal A and the terminal B are respectively connected via the SIP control function-equipped speech coding system conversion device T to perform communication. First, the functions and outlines of these units are described. explain.

  First, in the present embodiment, the speech control method conversion apparatus T with the SIP control function performs the speech control method of the terminals in SIP control VoIP communication between the terminals A and B whose bandwidth allocated during communication changes. Is provided with a function for preventing communication termination due to insufficient bandwidth by dynamically switching, mutually converting, and relaying.

  Note that, in the VoIP communication of this embodiment, the terminal A and the terminal B are based on SIP as a session start protocol for enabling mutual communication with the speech control method conversion apparatus T with the SIP control function. Accordingly, the speech control method conversion apparatus T with the SIP control function also has a function as a SIP proxy server for transferring the SIP message and a gateway function for transferring the SIP to a specific network.

  On the other hand, the terminal A includes one or a plurality of speech coding schemes, and performs communication with the speech coding scheme conversion apparatus T with a SIP control function via the speech codec A. Similarly, the terminal B also includes one or a plurality of speech coding schemes, which communicate with the speech coding scheme conversion apparatus T with the SIP control function via the speech codec B. However, the speech encoding methods provided in the terminal A and the terminal B do not need to be the same. In addition, as the terminals A and B, portable wireless terminals are assumed. Specifically, it may be a SIP telephone, a handheld computer, a palmtop computer, a notebook personal computer, a desktop personal computer, a workstation, or the like, or may be a SIP-compatible mobile phone, PHS (Personal Handyphone System), or the like. Furthermore, these terminals A and B function as SIP endpoints, and can interpret and process protocols and send and receive SIP messages. An arbitrary one of terminals used in various communication areas 14a to 14d having a VoIP communication function by SIP control is a terminal A, and a terminal used in various communication areas 14e to 14f having a VoIP communication function by SIP control is also used. Communication in a manner described in detail later is performed with any one as the terminal B.

  Here, each area such as various communication areas 14a to 14f having a VoIP communication function by SIP control functions as, for example, a wireless local area network (WLAN) communication network, a PHS communication network, a mobile phone communication network, or the like. . In addition, a plurality of service areas of each of these communication networks are mixed on the unified communication network 11 in which the mobile communication network and the heterogeneous communication network are connected and operated in an integrated manner.

Next, with reference to FIG. 2, a specific configuration of the above-described speech encoding method conversion apparatus T with the SIP control function will be described.
As shown in FIG. 2, the speech control method conversion apparatus T with the SIP control function includes a speech encoding method control unit 21, a SIP control unit 22, a speech encoding method conversion unit Xba23, and a speech encoding method conversion unit. The Xab 24, the bandwidth measuring unit 25, and the IP communication processing unit 26 are provided.

  Among these, the speech coding scheme control unit 21 determines a speech coding scheme that can be used in the terminal in communication based on a measurement result by bandwidth measurement or bandwidth information acquired by notification from the terminal, and the determined speech This is a part that performs control to switch or convert to the encoding system, in this example, the audio encoding system SCa or SCb.

  Further, the SIP control unit 22 has a SIP terminal control function and a proxy server function, and provides a speech code to the corresponding speech coding method conversion unit among the speech coding method conversion units described below. This is a part for exchanging a message corresponding to the content instructed by the system control unit 21 with the corresponding terminal via the IP communication processing unit 26.

  Also, the speech encoding scheme conversion unit Xba (first speech encoding scheme conversion unit) 23 is assigned to the terminal A with speech encoded data sent from the terminal B based on the control of the speech encoding scheme control unit 21. This is a part for conversion to a speech coding system SCa that can be used with a wide bandwidth.

  Similarly, the speech encoding scheme conversion unit Xab (second speech encoding scheme conversion unit) 24 allocates speech encoding data sent from the terminal A to the terminal B based on the control of the speech encoding scheme control unit 21. This is a part for conversion to a speech coding system SCb that can be used with a given bandwidth.

  The bandwidth measuring unit 25 is a part that measures the usable bandwidth allocated to each of the terminal A and the terminal B by, for example, a probing method. The bandwidth measurement by this probing method is detailed in, for example, “Tsuneo Katsuyama, Motomitsu Adachi“ Traffic Measurement Technology: NEPRI ”, FUJITSU.51, 6, p391-p395 (11, 2000)”. For this reason, the description which overlaps here is omitted, and only the outline | summary is demonstrated along the later communication procedure example.

The IP communication processing unit 26 is a part that performs IP communication between various terminals of the IP communication network including the terminal A and the terminal B.
FIG. 3 is a sequence chart showing a communication procedure between the terminal A and the terminal B centering on the communication control apparatus according to the present embodiment, that is, the speech coding system conversion apparatus T with the SIP control function. Will be described in detail with respect to an example of a communication mode between the terminals. Note that the procedure numbers (1) to (14) shown in FIG. 3 respectively correspond to the communication procedures (1) to (14) listed below. In the following description, for the sake of convenience, description of rewriting of headers of SIP and SDP (Session Description Protocol) messages that are normally processed is omitted.

Hereinafter, communication procedures are listed in order according to FIG.
(1) The terminal A sends an “INVITE (invite)” message to the terminal B to the speech encoding method conversion apparatus T with the SIP control function as the SIP proxy server. It is assumed that the “m =” and “a =” fields of the SDP portion of the message include some or all of the speech coding schemes that can be used by terminal A. Incidentally, the terminal A has two speech coding schemes “PCMA” and “G.729”, and the terminal B has two speech coding schemes “PCMU” and “GSM (GSM-FR)”. An example of these “m =” and “a =” fields in the case of Of course, direct communication between these terminals is not possible unless the speech coding scheme is converted by the speech coding scheme conversion apparatus T with SIP control function.

Terminal A (SDP part)

1: m = audio 4004 RTP / AVP 8
2: a = rtpmap: 8 PCMA / 8000
3: m = audio 4006 RTP / AVP 18
4: a = rtpmap: 18 G729 / 8000

Here, the first line represents an “m =” field, which indicates a media type “voice”, a reception port “4004”, a transport protocol type “RTP / AVP Profile”, and a payload type value “8”. The second line represents an “a =” field and represents an attribute related to the payload type “8” of “RTP / AVP Profile”. In this example, the codec “PCM A-law” and the sampling rate “8000 Hz” are specified together. Similarly to the first line, the third line represents the “m =” field, the media type “voice”, the reception port “4006”, the transport protocol type “RTP / AVP Profile”, the payload type value “ 18 ". The fourth line represents the “a =” field in the same manner as the second line, and represents an attribute related to the payload type “18” of “RTP / AVP Profile”. In addition, a codec (“G.729” in this example) and a sampling rate (“8000 Hz” in this example) are designated.

Terminal B (SDP part)

1: m = audio 24004 RTP / AVP 0
2: a = rtpmap: 0 PCMU / 8000
3: m = audio 24006 RTP / AVP 3
4: a = rtmap: 3 GSM / 8000

Here, the first line represents an “m =” field, and indicates a media type “voice”, a reception port “24004”, a transport protocol type “RTP / AVP Profile”, and a payload type value “0”. The second line represents an “a =” field and represents an attribute related to the payload type “0” of “RTP / AVP Profile”. In this example, the codec “PCM μ-law” and the sampling rate “8000 Hz” are specified together. Similarly to the first line, the third line represents the “m =” field, the media type “voice”, the receiving port “24006”, the transport protocol type “RTP / AVP Profile”, the payload type value “ 3 ". Similarly to the second line, the fourth line also represents an “a =” field, and represents an attribute related to the payload type “3” of “RTP / AVP Profile”. In addition, a codec (“GSM (GSM-FR)” in this example) and a sampling rate (“8000 Hz” in this example) are designated.

  In the description of the SDP section, each port of the “m =” field can be designated separately or collectively. At this time, the bandwidth allocated to the terminal A may be notified by the “b =” field.

  (2) The SIP control unit 22 deletes the “Via (via):” header from the terminal A to the speech encoding method conversion apparatus with SIP control function T for the SIP part of the message of the procedure (1). The IP address or domain name part (after @) of the “Contact:” header is rewritten to the IP address assigned to the IP communication processing unit 26. Then, the “o =” and “c =” fields of the SDP unit are also rewritten to the IP address assigned to the IP communication processing unit 26, and the port number in the “m =” field is rewritten to the number assigned to the media session for the terminal B. . In addition, the “m =” field and the “a =” field are rewritten to a list of speech encoding methods that can be converted by each speech encoding method conversion unit. Furthermore, the IP communication processing unit 26 is instructed to open a port for the assigned part number. Also, when there is bandwidth information in the “b =” field, this is notified to the speech coding scheme control unit 21. Thereafter, the IP communication processing unit 26 sends the message thus processed to the terminal B.

  (3) The bandwidth measurement unit 25 starts measurement using the probing method for the bandwidth that can be used by the terminal A and the terminal B. In this measurement, first, a measurement packet is transmitted to a communication path of the network at a constant speed, and its round trip delay (RTT) is measured. When a predetermined strong correlation is detected between the RTTs of two consecutive measurement packets, the bandwidth that can be used by the terminal A and the terminal B is obtained from the transmission speed of the measurement packet at that time. The measurement packet size at the start of measurement is the same size as that of the speech encoding method having the largest packet size among the speech encoding methods notified by the message of the procedure (1). Then, every time a significantly large delay is detected, the size is sequentially switched to the same size as that of the voice encoding method having a smaller packet size. The bandwidth is measured intermittently at regular intervals, and the measurement result is notified to the speech coding scheme control unit 21.

  Note that if the measurement packet is transmitted in addition to the voice encoded packet during voice communication, the voice delay causes a shortage of bandwidth due to the measurement packet when the bandwidth is marginal, resulting in trouble with voice communication. May cause. In such a case, the transmission of the measurement packet is interrupted, and the jitter (the difference between the arrival time of the packet and the time stamp embedded in the packet (in the RTP header)) or the packet discard rate is measured. By doing so, the operation of the bandwidth measuring unit can be switched so as to monitor the line state. If the transmission frequency of measurement packets is properly managed, jitter, delay, etc. that adversely affect voice quality can be minimized, but once the original communication path is very thin, it occurs once. This is an operation for relieving such a situation since it may take a long time to eliminate the jitter and the like, and the voice quality may be affected during that time.

  In addition, since encoded speech packets are transmitted at regular intervals (for example, every 20 ms), if the bandwidth is insufficient, packets accumulate at the bottleneck, delaying the arrival time of packets, and increasing jitter. To go. If the jitter becomes so large that it cannot be solved by a jitter buffer or the like, the sound is finally reproduced with a large delay at which the sound is interrupted. Therefore, it can also be determined that a shortage of bandwidth occurs when the jitter exceeds the allowable value.

  Packets accumulated in the bottleneck may be discarded as processing such as congestion control for routers, etc., but since each packet is assigned a sequence number, the discarded packets can be detected. . That is, even when the packet discard rate that hinders voice communication is reached, it can be used as a trigger for switching the voice coding system because the bandwidth is insufficient.

  Furthermore, a measurement method that does not impose a bandwidth by combining active measurement of transmitting a main measurement packet and passive measurement of monitoring the reception state of an auxiliary voice packet is possible. In this case, when a voice packet is not transmitted, it is naturally necessary to transmit a measurement packet.

  (4) The speech coding scheme control unit 21 determines a speech coding scheme SCa that can be used by the terminal A based on the measurement result of the bandwidth measuring unit 25 or information notified from the terminal, and uses this as the speech coding scheme. The conversion unit Xba23 is instructed. At the same time, the bandwidth measuring unit 25 uses a measurement packet having the same size as that of the voice encoding scheme SCa for the terminal A.

  (5) In response to the message of the procedure (2), the terminal B returns a “180 Ringing (provisional response)” message. Subsequently, one of the speech encoding methods indicated in the “m =” and “a =” fields rewritten in the procedure (2) is selected, and the selection information is written in the “m =” field. . In addition, for the “a =” field, information of the speech coding scheme related to the selection (further, a list of part or all of speech coding schemes that can be used by the terminal B may be added. In this case, “m =” The port information is also added to the field), and a “200 OK (success response)” message is returned to the speech control method conversion apparatus T with the SIP control function. At this time, the bandwidth allocated to the terminal B may be notified to the speech encoding method conversion apparatus with SIP control function T by the “b =” field.

  (6) The SIP control unit 22 adds the “Via:” header from the terminal A deleted in the procedure (2) to the “180 Ringing” message returned from the terminal B in the procedure (5). The return process to the terminal A is performed. In addition, regarding the “200 OK” message, the “Contact” is sent so that the terminal A sends not the terminal B but the “ACK (confirmation of final response to INVITE)” message to the speech coding method conversion apparatus T with SIP control function. : "The IP address or domain name part (after @) of the header is rewritten to the IP address assigned to the IP communication processing unit 26. Then, the “Via:” header from the terminal A deleted in the procedure (2) is added to the “200 OK” message. Further, the SIP control unit 22 assigns the “o =” and “c =” fields of the SDP unit to the IP address assigned to the IP communication processing unit 26 and the port number of the “m =” field to the media session for the terminal A. Rewrite the number. In addition, the “m =” and “a =” fields are rewritten to the speech coding scheme (SCa) determined by the speech coding scheme control unit 21 in the procedure (4), and the terminal in the original “m =” field is rewritten. The voice encoding scheme SCb selected by B is notified to the voice encoding scheme control unit 21. Thereafter, the SIP control unit 22 instructs the IP communication processing unit 26 to open a port for the assigned port number. As a result, the IP communication processing unit 26 sends the processed message to the terminal A. Further, the bandwidth measuring unit 25 uses a measurement packet having the same size as that of the voice encoding system SCb for the terminal B thereafter.

  (7) The voice coding method control unit 21 instructs the voice coding method conversion unit Xab24 of the content according to the voice coding method notified in the procedure (6). As a result, the speech encoding system conversion unit Xab24 is set to perform conversion from SCa to SCb. Note that the other speech coding method conversion unit Xba23 is set to perform conversion from SCb to SCa in the previous procedure (4).

(8) In response to the message of the procedure (6), the terminal A transmits an “ACK” message to the speech control method conversion apparatus T with the SIP control function.
(9) The SIP control unit 22 deletes the “Via:” header from the terminal A to the speech encoding method conversion apparatus with SIP control function T for the SIP part of the message of the procedure (8). Then, the IP communication processing unit 26 transmits the processed message including the “ACK” message to the terminal B. The “INVITE” message requires a three-way handshake of “INVITE”, “200 OK”, and “ACK”, and another “INVITE” method is issued until one “INVITE” method is completed. Can not do it. For this reason, in the example of FIG. 3, terminal A completes one method through procedure (1) -procedure (6) -procedure (8), and terminal B performs procedure (2) -procedure (5)- One method is completed through step (9).

  (10) Upon reception of each of the “ACK” messages, the terminal A and the terminal B follow the “SDP” message rewritten by the speech coding method conversion apparatus T with the SIP control function, not directly between the terminals, respectively. A media session is established with the speech encoding method conversion apparatus T with a control function. As a result, the terminal A and the speech encoding scheme conversion apparatus T with the SIP control function have the speech encoding scheme SCa, and the terminal B and the speech encoding scheme conversion apparatus with the SIP control function T have the speech encoding scheme SCb. It becomes the used VoIP communication.

  (11) The speech coding scheme control unit 21 determines that the bandwidth measured by the bandwidth measuring unit 25 is equal to or less than the bandwidth necessary for the speech coding scheme currently used by either the terminal A or the terminal B. When it becomes or is likely to happen, the following control is further executed. That is, in such a case, the speech coding scheme control unit 21 applies a “re-INVITE” message for switching to a lower bit rate speech coding scheme to the SIP control unit 22. Instruct to issue only for the terminal. Note that the speech coding schemes indicated here are in the speech coding scheme list in the “m =” or “a =” field of the message exchanged in the procedure (1) or the procedure (5), This is one of speech coding schemes that can be used with the measured bandwidth. In addition, the speech coding method of the terminal on the side where the bandwidth does not change is left as it is. However, if not all speech coding schemes can be used for a terminal that needs to be switched, it is not a switching of the speech coding scheme, but a pending message (the “c =” field in the SDP section is “c = IN IP40”). .0.0. ”Is sent to both terminals to instruct to temporarily stop VoIP communication.

  Note that the “re-INVITE” message is an “INVITE” message including a parameter for switching the speech encoding method. This “re-INVITE” message is sent to the terminal A and the terminal B in the procedure (1) -procedure (6) -procedure (8) or the procedure (2) -procedure (5) -procedure (9). It is generated with reference to the information exchanged. In the procedure (10) and subsequent steps, the “re-INVITE” message and the like are exchanged independently for each of the terminals A and B. Therefore, the management of the “Cseq (command sequence):” header and the like is performed for each terminal. . However, since it is necessary to match the sequence number of the “Cseq:” header in both terminals such as the “BYE (session end)” method, the management is assumed to be coordinated.

  Further, the above-mentioned pending message is one of “re-INVITE” messages, but does not change the voice encoding method. Therefore, the “m =” and “a =” fields indicate values currently in use. As it is, “c = IN IP4 0.0.0.0” is set. Then, the hold message is transmitted to each of the terminal A and the terminal B as described above. Here, “hold” means to stop the transmission of the VoIP packet while maintaining the communication (session).

  (12) The SIP control unit 22 generates a “re-INVITE” message or a hold message for the terminal specified in the procedure (11). Then, the IP communication processing unit 26 sends this message to the corresponding terminal.

  (13) When the SIP control unit 22 receives the “200 OK” message from the corresponding terminal in the “re-INVITE” sequence for switching the speech encoding method, the SIP control unit 22 informs the speech encoding method control unit 21 to that effect. Notice.

  (14) When the above notification is received from the SIP control unit 22, the speech encoding method control unit 21 sends the corresponding speech encoding method conversion unit to switch from the original speech encoding method to the new speech encoding method. Instruct.

Through such a procedure, VoIP communication by SIP control is executed, and the VoIP communication is terminated after receiving a “BYE” method (not shown) from any terminal.
Next, features of the embodiment configured as described above will be described below.

  (A) In this embodiment, at the time of IP voice signal communication (VoIP communication) using a session start protocol (SIP) between terminals having one or a plurality of voice encoding schemes, they are assigned to the terminals in communication. Bandwidth is measured. And among these terminals, the terminals having a plurality of speech coding schemes are dynamically switched in accordance with the measured bandwidth. For this reason, even if the bandwidth allocated to the terminal in communication changes, the termination of communication due to such a change in bandwidth can be prevented, and it can be matched to the line conditions in each case. An appropriate speech encoding method can be selected. Specifically, by switching the “Contact:”, “Via:” header, etc., the speech coding method conversion apparatus T with SIP control function is always interposed, so that the terminal A and the terminal B cannot communicate directly. I have to. As a result, media sessions are established separately between the terminal A and the terminal B but not with the speech control method conversion apparatus T with the SIP control function. For this reason, the voice coding scheme conversion apparatus T with SIP control function assigns the bandwidths of the smallest required bandwidth among the voice coding schemes of the terminal A and the terminal B as Ca and Cb, respectively, to the terminals. If the existing bandwidths are Ba and Bb, respectively, communication is possible if the following conditional expression (2) is true (assuming logical product operation). That is, the conditional expression (1) capable of communication in the conventional method is equivalent to the following conditional expression (1 ′), but in comparison with the conditional expression (1 ′), the following conditional expression (3) or (4) Thus, communication can be maintained even when the bandwidth of the speech coding scheme that can be used on one side is larger than the bandwidth allocated to the other side. As a result, communication that is not easily affected by a change in the bandwidth of the communication partner and is not easily disconnected is realized. In addition, if the hold state is included, communication that is never disconnected can be realized.

(Ca ≦ Ba) AND (Cb ≦ Bb) (2)

(Ca ≦ Ba) AND (Cb ≦ Bb) AND (Ca ≦ Bb)
AND (Cb ≦ Ba) (1 ′)
(Ca ≦ Ba) AND (Cb ≦ Bb) AND (Cb> Ba) (3)
(Ca ≦ Ba) AND (Cb ≦ Bb) AND (Ca> Bb) (4)

(B) In the present embodiment, the bandwidth is measured based on the measurement of the round trip delay of the measurement packet transmitted to the communication path at regular time intervals. In addition, at this time, the bandwidth that can be used as the bandwidth is obtained from the transmission speed of the measurement packets when a predetermined correlation is detected in the round-trip delay between two consecutive measurement packets. However, the bandwidth measurement with high accuracy according to the line condition is performed each time.

  (C) In the present embodiment, when the bandwidth is measured, the measurement packet size at the start is the largest packet among the speech coding schemes notified in advance from a terminal having the plurality of speech coding schemes. The size is the same as that of the audio coding method. Each time a large delay exceeding a predetermined value is detected with respect to the round-trip delay of the measurement packet, the measurement packet is sequentially switched to the same size as that of the voice encoding method having a smaller packet size. For this reason, the above-described bandwidth measurement is performed more efficiently, that is, in a shorter time, and as a result, the dynamic switching of the speech coding scheme is performed more quickly and appropriately.

  (D) In this embodiment, based on the measurement of the bandwidth, at least one of the terminals lacks available bandwidth with respect to the bandwidth required for the speech coding scheme being used between the terminals. When the situation is detected, the voice coding schemes are dynamically switched in a manner of switching to a lower bit rate voice coding scheme. As a result, the termination of communication due to the aforementioned insufficient bandwidth is suitably suppressed.

  (E) In the present embodiment, when there is no speech coding method to be switched to the terminal that requires switching of the speech coding method, communication between the corresponding terminals is temporarily suspended. As a result, it is possible to wait for the resumption of communication based on the bandwidth recovery without performing a troublesome operation such as the reconnection operation described above.

  (F) The communication control device according to the present embodiment has a bandwidth measuring unit 25 in a conventional speech coding method conversion device with a SIP control function, and a function for dynamically switching the speech coding method based on the measurement result. It has been added. Therefore, application to VoIP communication by conventional SIP control is also easy.

In addition, the said embodiment can also be changed and implemented as follows.
In the above embodiment, the hold message is a message in which the “c =” field of the SDP part becomes “c = IN IP4 0.0.0.0”. However, as the hold message, “c = IN IP4 0.0. For example, “a = sendly” may be added to the SDP unit. Even in this case, the same effect as that of the above embodiment can be obtained.

  In the above embodiment, when the bandwidth measured by the bandwidth measuring unit 25 is or is likely to be less than the bandwidth required for the speech coding scheme that the terminal is currently using, the lower bit rate It is instructed to issue only the “re-INVITE” message for switching to the speech encoding method for the terminal. Instead, the terminal that has detected that a problem has occurred in the bandwidth itself sends a “re-INVITE” message to the speech encoding method conversion apparatus T with the SIP control function, and switches the speech encoding method. You may request. Further, the hold message can also be sent from the terminal side to the speech encoding method conversion apparatus T with the SIP control function. In this case, in the speech coding method conversion apparatus T with the SIP control function, the hold message is transferred to the other terminal after rewriting the header in the same manner as in the procedure (2), and both the terminals are put on hold.

  In the above embodiment, the measurement packet size at the start of measurement is the same size as that of the speech encoding method having the largest packet size among the speech encoding methods notified by the message of procedure (1). When a significantly large delay is detected, the size is sequentially switched to the same size as that of the voice encoding method having a smaller packet size. However, when the usable bandwidth increases and a speech coding method with a higher bit rate and better sound quality can be used, the speech coding method control unit 21 determines whether or not the speech coding method control unit 21 based on the measurement result of the bandwidth measurement unit 25. Alternatively, the terminal itself may switch the speech encoding method by a “re-INVITE” message. As a result, when the bandwidth increases, the bit rate can be increased and communication can be performed with better sound quality. Similarly, when the original speech encoding method becomes usable during the hold, the VoIP communication can be resumed by the “hold release” message.

-The bandwidth measurement method is not limited to the one exemplified above, for example,
(A) A method for measuring the reachability of a host on a route and the RTT (Round Trip Time) of a packet by ping using an ICMP (Internet Message Protocol) echo packet.
(B) By sending UDP (User Datagram Protocol) echo packets while increasing TTL (Time To Live) one by one, the IP address of the router on the route is checked one by one, and any router existing on the route Method to measure the RTT and the discard rate.
(C) Probing by combining UDP packets of various sizes, obtaining a minimum value of RTT, and using a tool for measuring the bandwidth of each hop link from the increasing tendency of the RTT minimum value relative to the packet size (pathchar) ).
(D) A method of transmitting a packet while performing a congestion avoidance algorithm while measuring a link performance by receiving a UDP TTL excess packet and transmitting a UDP echo packet while increasing TTL as in the above “traceroute” ( Treno).
Etc. can be adopted as appropriate. However, the above (d) “Treno” and the above (c) “pathchar” transmit a large number of packets in order to improve the measurement accuracy and put a load on the network, so that there is a limitation in practical use.

  In the above embodiment, both the terminals A and B are provided with one or a plurality of speech encoding methods. However, in order to perform VoIP communication more efficiently, both of these terminals are provided with a plurality of speech encoding methods. It is desirable that

  In the above embodiment, portable wireless terminals are assumed as the terminal A and the terminal B. However, the terminals A and B are not limited to portable wireless terminals, and wired terminals connected to a wired communication network are used. You can also Even in this case, the same effect as that of the above embodiment can be obtained.

  In the embodiment described above, the SIP is used as the session start protocol. However, the session start protocol is not limited to the SIP, but includes ITU-T recommendation (standard) “H.323”. Even in this case, it is possible to realize communication control in a mode according to the above embodiment.

The block diagram for demonstrating the schematic structure about the unified communication network with which one Embodiment of this invention is applied. The block diagram which shows the specific structural example of the audio | voice coding system converter with a SIP control function. The sequence chart for demonstrating the VoIP communication execution procedure of this embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Unified communication network, 14a-14f ... Various communication area which has VoIP communication function by SIP control, 21 ... Voice coding system control part, 22 ... SIP control part, 23 ... Voice coding system conversion part Xba, 24 ... Voice Coding method conversion unit Xab, 25... Bandwidth measurement unit, 26... IP communication processing unit, T.

Claims (10)

In a communication control method for controlling IP audio signal communication using a session start protocol between terminals having one or a plurality of audio encoding methods,
While measuring the bandwidth allocated to the terminal in communication, among those terminals, a terminal having a plurality of voice encoding schemes is dynamically changed in accordance with the measured bandwidth. A communication control method characterized by switching. The bandwidth measurement is performed by transmitting measurement packets to the communication path of the communicating terminal at a predetermined time interval, measuring the round-trip delay, and calculating a predetermined round-trip delay between two consecutive measurement packets. The communication control method according to claim 1, wherein the communication control method is performed by obtaining a bandwidth that can be used as the bandwidth from a transmission rate of the measurement packets when the correlation is detected. The measurement packet size at the start of measurement of the bandwidth is the same size as that of the speech encoding method having the largest packet size among speech encoding methods notified in advance from a terminal having the plurality of speech encoding methods. Whenever a large delay exceeding a predetermined value is detected with respect to the round-trip delay of the measurement packet, the measurement packet is sequentially switched to the same size as that of the voice encoding method having a smaller packet size. The communication control method according to 2. Based on the measurement of the bandwidth, when a situation is detected in which at least one terminal lacks available bandwidth relative to the bandwidth required for the speech coding scheme in use between the terminals, The communication control method according to any one of claims 1 to 3, wherein the dynamic switching of the voice encoding method is performed in a mode of switching to a lower bit rate voice encoding method. The communication control method according to claim 4, wherein when a voice encoding method to be switched does not exist in a terminal that needs to be switched, the communication between the corresponding terminals is temporarily suspended. A communication control apparatus for controlling IP audio signal communication using a session start protocol between terminals having one or a plurality of audio encoding methods,
First and second speech coding scheme conversion units each having a function of transforming speech coding data sent from one of the communicating terminals into a speech coding scheme that can be used with a bandwidth allocated to the other terminal When,
A bandwidth measuring unit that measures the available bandwidth allocated to each terminal in communication;
An IP communication processing unit for performing IP communication processing between various terminals;
The first and second speech codes are determined so as to determine a speech encoding method that can be used in the communicating terminal based on a bandwidth measurement result by the bandwidth measuring unit, and to convert to the determined speech encoding method. A voice coding method control unit that instructs the corresponding voice coding method conversion unit among the coding method conversion units;
A message having a function of controlling a terminal based on the session start protocol and a proxy server function, and sending a message corresponding to the content instructed from the speech coding method control unit to the corresponding speech coding method conversion unit in the IP communication A SIP control unit that exchanges data with a corresponding terminal via the processing unit;
A communication control apparatus comprising: The bandwidth measuring unit transmits measurement packets to the communication path of the communicating terminal via the IP communication processing unit at a constant time interval, measures the round-trip delay, and performs two consecutive measurement purposes. The communication control device according to claim 6, wherein a bandwidth that can be used as the bandwidth is obtained from a transmission rate of the measurement packets when a predetermined correlation is detected in the round-trip delay of the packet. The bandwidth measuring unit has the largest packet size of speech encoding methods notified in advance from a terminal having the plurality of speech encoding methods as the size of the measurement packet at the start of the bandwidth measurement. Whenever a large delay exceeding a predetermined value is detected with respect to the round-trip delay of the measurement packet, the packet size is the same as that of the encoding method. The communication control device according to claim 7, wherein the size is sequentially switched to a packet size. Based on the measurement of the bandwidth by the bandwidth measurement unit, the speech coding method control unit has a usable bandwidth for a bandwidth required for the speech coding method being used between the terminals. The method according to any one of claims 6 to 8, wherein when a shortage is detected in at least one of the terminals, the speech coding scheme is switched to a speech coding scheme having a lower bit rate. Communication control device. The speech coding scheme control unit sends a message to temporarily hold communication between the corresponding terminals when the speech coding scheme to be switched does not exist in the terminal that needs to switch the speech coding scheme. The communication control device according to claim 9, wherein the communication control device transmits the data to the terminals via the SIP control unit and the IP communication processing unit.

Images