JP2014072577A - Voice application system, management server device, and control method thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce load in the whole system when voice application processing is performed.SOLUTION: The voice application system has a media server device, an application control server device, and a management server device. The management server device has storage means and processing means. The storage means creates a management table in which a session ID for specifying a session and a server ID for specifying the application control server device and the media server device respectively in which a session is established are associated with each other and stores the management table. The processing means performs processing using a voice application to a request source for each session, on the basis of the management table stored by the storage means.

Description

  Embodiments of the present invention relate to, for example, a SIP (Session Initiation Protocol) server, a MRCP (Media Resource Control Protocol) server, and a voice application system in which a media server executes voice application processing related to an exchange service, a management server device, and a control method thereof .

  In offices and business offices, a private branch exchange (PBX) or a button telephone device is generally used as a device for connecting an extension telephone network and a public telephone network. These telephone exchange devices are configured as dedicated devices for exchanging and controlling voice lines.

  On the other hand, a system has also been proposed in which a server device is connected to a telephone exchange device, and this server device executes a part of the exchange connection processing of the telephone exchange device. The server equipment includes SIP (Session Initiation Protocol: RFC 3261), MRCPv2 (Media Resource Control Protocol version 2: draft-ietf-speechsc-mrcpv2-27), RTP (Real-Time Transfer Protocol: RFC) A speech synthesis server controlled by 1889) or a speech application server that is a speech recognition server is used.

JP 2007-60381 A

  By the way, in the above-mentioned voice application server, the voice application server receives a request from the client and performs a SIP control when executing the process, MRCP controlled as an application by SIP control, and controlled as a media resource RTP to be placed in the same device. For this reason, it is necessary for the voice application server to execute MRCP as an application module and RTP as a media resource in the same device in response to a request from a client, and the processing tends to be overloaded as a device. is there.

  An object of the present invention is to provide a voice application system, a management server device, and a control method thereof that can reduce the load on the entire system when executing voice application processing.

  According to the embodiment, the voice application system includes a media server device, an application control server device, and a management server device. The media server device executes a predetermined voice application. The application control server device is connected to the media server device via an IP (Internet Protocol) network, and controls the execution of the voice application. The management server device is connected to the media server device and the application control server device via an IP (Internet Protocol) network, and can connect a telephone exchange device that performs call connection between a plurality of telephone terminals. Controls a session for communication connection between control server devices. The management server device includes a storage unit and a processing unit. The storage means, when a session for communication connection is established between the usable media server device and the application control server device based on at least a voice application execution request from the telephone terminal, A management table is created and stored in association with the server ID that identifies each of the media server device and the application control server device with which the session is established. The processing means performs processing using the voice application for the request source for each session based on the management table stored in the storage means.

1 is a block diagram showing a configuration of a voice application system according to an embodiment. The block diagram which shows the function structure of the SIP server which concerns on one Embodiment. The figure which shows an example of the memory content of MRCP server DB and media server DB which concerns on one Embodiment. The figure which shows an example of the memory content of session management DB which concerns on one Embodiment. The block diagram which shows the function structure of the MRCP server which concerns on one Embodiment. The block diagram which shows the function structure of the media server which concerns on one Embodiment. The figure which shows an example of the table used when performing a voice dial service. The schematic block diagram which shows the comparative example of a voice application system. The sequence diagram which shows the transmission / reception operation | movement of the signal between the telephone terminal at the time of requesting the voice application service in this embodiment, a telephone switching apparatus, IVR, a SIP server, a MRCP server, and a media server. The sequence diagram which shows the control procedure of the telephone exchange apparatus, IVR, and a media server at the time of performing the speech synthesis process and speech recognition process in this embodiment.

  Hereinafter, embodiments will be described with reference to the drawings.

  FIG. 1 is a block diagram illustrating a configuration of a voice application system according to an embodiment. In this voice application system, the voice application server 10 is distributed for each module to the SIP server 11, the MRCP servers 12, 13 and the media servers 14, 15, and each SIP server 11, the MRCP servers 12, 13 and the media server 14, 15 are connected to each other by the IP network INW. The SIP server 11 is connected to a telephone exchange device 30 via an automatic voice response device (IVR) 20.

  The telephone exchange device 30 accommodates a plurality of telephone terminals 41 to 4n through extensions, and transmits incoming signals to the telephone terminals 41 to 4n, detects outgoing signals from the telephone terminals 41 to 4n, and the like. An extension interface process is performed for each of the telephone terminals 41 to 4n. The telephone exchange device 30 can be connected to the analog trunk 50, and performs trunk interface processing for the analog trunk 50 such as transmission of a transmission signal to the analog trunk 50 and detection of an incoming call from the analog trunk 50. Further, the telephone exchange device 30 connects the telephone terminals 41 to 4n, the telephone terminals 41 to 4n and the analog trunk 50, and the telephone terminals 41 to 4n and the IVR 20 to connect these terminals. Enables transmission of audio signals between the two.

  For example, a call from the telephone terminal 41 to the telephone exchange device 30 is transferred to the IVR 20 that is a voice application client. The IVR 20 responds to incoming calls, automatically connects to speech synthesis (TTS: Text To Speech) and speech recognition (ASR: Auto Speech Recognition) applications, and dynamically responds to customer terminals. It is possible to provide a system.

  In the present embodiment, session generation is performed by the SIP server 11 using SIP as a protocol, and speech synthesis requests and speech recognition requests are performed by the MRCP servers 12 and 13 using MRCPv2 as a protocol. The media servers 14 and 15 perform voice recognition processing and RTP transmission / reception. In the present embodiment, two MRCP servers and two media servers are configured, but the number of servers is not particularly limited due to the configuration.

FIG. 2 is a block diagram showing a functional configuration of the SIP server 11.
In addition to the SIP control unit 111 that controls session management by SIP, the SIP server 11 includes a server monitoring control unit 112 that performs monitoring control of the MRCP servers 12 and 13 and the media servers 14 and 15 and an MRCP server DB that stores monitoring status. A (database) 113 and a media server DB 114 are installed. Further, the SIP server 11 includes a communication unit 115, an interprocess communication control unit 116, and a session management DB 117.

  The communication unit 115 performs interface processing for transferring control signals and audio signals to and from the IP network INW.

  As shown in FIG. 3, the MRCP server DB 113 and the media server DB 114 store a table for monitoring the state of the MRCP server and the media server. The server monitoring control unit 112 acquires setting information (server ID, IP address) of the MRCP servers 12 and 13 in the SIP server 11 and monitors the state of the MRCP servers 12 and 13. In the present embodiment, as a method for monitoring the state of the MRCP servers 12 and 13, a method such as mutual communication using an arbitrary protocol or mutual communication using Ping is used, but the method is limited in the present embodiment. It is not a thing.

  Then, the status monitoring result of the MRCP server is recorded in the table as “Active” or “Stopped” in real time, and stored in the MRCP server DB 113. Similarly, in the media servers 14 and 15, the server monitoring control unit 112 performs status monitoring, and the status monitoring result is stored in the media server DB 114.

  When the voice application execution request arrives from the IVR 20, the SIP control unit 111 is used for communication connection between the MRCP server 12 and the media server 14 that can be used based on the stored contents of the MRCP server DB 113 and the media server DB 114. When a session is established, a session management table in which a session ID that identifies the session and a server ID that identifies each of the MRCP server 12 and the media server 14 with which the session is established is created and stored in the session management DB 117. .

  The session management DB 117 stores a session management table in which IDs for identifying MRCP servers and media servers corresponding to session IDs are stored as shown in FIG. The SIP control unit 111 manages the MRCP server and media server information allocated to the session generated by the request from the IVR that is the voice application client by using the session management DB 117. In the present embodiment, a round robin method, a weighted round robin method, a minimum connection method, or the like is used as a method for allocating processes to the MRCP server and the media server. However, the method is limited in the present embodiment. It is not something.

  The inter-process communication control unit 116 is a module that controls communication between the SIP server, the MRCP server, and the media server. The inter-process communication control unit 116 performs mutual communication with each server via the communication unit 115, and also from the IVR 20 that is a voice application client. It also communicates voice application requests. When an audio signal is sent from the request source, the session management table stored in the session management DB 117 is referred to, and based on the reference result, the audio by the media servers 14 and 15 is sent to the request source for each session. Provide services using applications.

FIG. 5 is a block diagram showing a functional configuration of the MRCP servers 12 and 13. Here, the MRCP server 12 will be described as a representative.
The MRCP server 12 includes a communication unit 121, an MRCP control unit 122, and an interprocess communication control unit 123. Note that the communication unit 121 and the inter-process communication control unit 123 are the same as the communication unit 115 and the inter-process communication control unit 116 of the SIP server 11, and thus detailed description thereof is omitted. The MRCP control unit 122 executes processing of an MRCP request from the IVR 20 that is a voice application client via the communication unit 121.

FIG. 6 is a block diagram showing a functional configuration of the media servers 14 and 15. Here, the media server 14 will be described as a representative.
The media server 14 includes a communication unit 141, an RTP control unit 142, an inter-process communication control unit 143, a voice application control unit 144, and a voice application engine unit 145. Note that the communication unit 141 and the inter-process communication control unit 143 are the same as the communication unit 115 and the inter-process communication control unit 116 of the SIP server 11, and detailed description thereof is omitted. The RTP control unit 142 performs control for performing RTP transmission / reception with the IVR 20 that is the voice application client via the communication unit 141.

  The voice application control unit 144 is connected to the storage unit 146 and controls the voice application engine unit 145 and the storage unit 146. The voice application control unit 144 analyzes the contents of the voice synthesis request and the voice recognition request from the MRCP servers 12 and 13, The voice application engine unit 145 performs voice synthesis processing and voice recognition processing. Further, the voice application control unit 144 can notify the IVR 20 of the character string transmitted from the voice recognition unit 1451 of the voice application engine unit 145 via the MRCP server 12 or the MRCP server 13, for example.

  The IVR 20 has, for example, a table for associating the name of the call destination shown in FIG. 7 with the telephone number of the user, and this table is used when performing a voice dial service.

  Here, the voice dialing service is a method in which a name and a telephone number of a callee are registered in advance, and a voice recognition unit 1451 converts the callee name into a character string when the caller inputs the name of the callee. The corresponding telephone number is read from the above table, and the telephone number is notified to the telephone exchange device 30 and transmitted.

  The voice recognizing unit 1451 has a voice-character string conversion table (not shown). The voice recognition unit 1451 receives a voice signal representing the name of a call destination sent from one of the telephone terminals 41 to 4n when the voice dial service is executed. And, it is received via the MRCP server 12 or the MRCP server 13 and converted from the waveform of the voice signal into a character string using the above table. In addition, the voice recognition unit 1451 notifies the voice application control unit 144 of the character string after conversion.

  Further, the voice application engine unit 145 includes a voice synthesis unit 1452. The voice synthesizing unit 1452 generates a predetermined voice guidance based on the processing result of the MRCP server 12 or the MRCP server 13 and executes one of the telephone terminals 41 to 4n which are calling telephone terminals when executing the voice dial service process. It is a notification. The content of the voice guidance generated by the voice synthesizer 1452 is “Please input the name of the destination” when the MRCP server 12 or the MRCP server 13 instructs the start of the voice dial service process. When there is an instruction that an object has been uttered again but could not be recognized, there was a message such as "Could not recognize. Please speak slowly again."

  FIG. 8 is a schematic configuration diagram showing a conventional voice application system. As shown in FIG. 8, when there are calls from the plurality of telephone terminals 41 and 42 to the telephone switching apparatus 30 at the same time and the response is executed by the IVR 20, the IVR 20 serves as a SIP client with the voice application system 10 as a voice application. When processing is dynamically requested, the IVR 20 has transmitted a SIP INVITE message, which is a request to start a session, to the voice application system 10, and the voice application system includes a SIP server module, an MRCP server module, and a media server module. In the voice application system 10, the MRCP request that is a subsequent request from the IVR 20 and the RTP transmission / reception that is a medium must all be performed in the same device. However, the load on the device may increase due to simultaneous processing of SIP / MRCP / voice application processing / media transmission / reception.

  Even when a plurality of voice application requests arrive, the voice application processing in the voice application system 10 must be simultaneously processed in the apparatus, and the processing load may increase when the requests are congested. there were.

  In this embodiment, in a telephone system including a private branch exchange using a voice application system and an automatic voice response device, the device is divided into modules of the voice application system 10, session management is performed by a SIP server, and application management is performed by an MRCP server. The media server performs application processing and media resource management, so that processing of a single voice application request is shared by a plurality of devices and performed manually, thereby reducing the overall processing of the voice application system 10. In addition, when a plurality of customers receive calls simultaneously on the telephone system, the automatic answering apparatus dynamically generates a plurality of voice application requests. The present invention proposes a processing distribution method aimed at reducing the processing load of the application system 10 and improving the input processing and output processing of audio data that is the result of the audio application processing. Will be described below with reference to FIGS.

  FIG. 9 shows transmission and reception of signals among the telephone terminal 41, the telephone exchange device 30, the IVR 20, the SIP server 11, the MRCP server 12, and the media server 14 when requesting the voice application service in the present embodiment. It is a sequence diagram which shows operation | movement.

  For example, as a request for request A shown in FIG. 1, after the user of the telephone terminal 41 goes off-hook, a function number for executing the voice dialing function is input via the keypad or function key (step 91a). A call request is notified to the telephone exchange device 30 together with the signal including the input function number signal (FIG. 9 (1)).

  Upon receiving the call request from the telephone terminal 41, the telephone exchange device 30 performs incoming call transfer processing (step 92a) and transfers the call request to the IVR 20 (FIG. 9 (2)).

  The IVR 20 is a device equipped with SIP and MRCPv2, and generates a new session for the SIP server 11 when voice application processing is required to respond to the request A, which is a request specified by the function number input from the telephone terminal 41. Request is made by SIP INVITE (step 93a and FIG. 9 (3)).

  The SIP server 11 receives a request for SIP session generation at the communication unit 115 (step 94a), and selects a server in the “in operation” state from the MRCP server DB 113 managed by the server monitoring control unit 112 at the SIP control unit 11. If the MRCP server 12 is available after the search (Yes in step 94b), the use of the MRCP server 12 is determined, and a session corresponding to the request A requested from the telephone terminal 41 is determined in the session management DB 117. The server ID “1” of the MRCP server 12 is registered in association with the session ID “1” (step 94c). If there is no MRCP server that can be used (No in step 94b), an error response is notified to the IVR 20 that is the client of the voice application.

  Then, the SIP server 11 notifies the MRCP server 12 of the start of the session via the interprocess communication control unit 116 ((4) in FIG. 9). Then, the MRCP server 12 acquires MRCP request information generated by the IVR 20 that is the voice application client notified from the SIP server 11 by the MRCP control unit 122 by starting a new session received via the inter-process communication control 123. Preparation for receiving an MRCP request from the IVR 20 is performed (step 95a).

  Subsequently, the SIP server 11 searches the media server DB 114 for a server that is in the “running” state, and if the media server 14 is usable (Yes in step 94d), the media server 14 is determined by an arbitrary determination method. In the session management DB 117, the server ID “2” of the media server 14 is registered in the session management DB 117 in association with the session ID “1” corresponding to the request A requested from the telephone terminal 41 ( Step 94e). If there is no usable media server (No in step 94d), an error response is notified to the voice application client IVR 20.

  Thereafter, the SIP server 11 notifies the media server 14 of the start of a session via the interprocess communication control 116 (FIG. 9 (5)).

  The media server 14 opens an RTP reception port from the IVR 20 that is a voice application client notified from the SIP server 11 by the RTP control unit 142 by starting a new session received via the inter-process communication control unit 143 (step S40). At the same time, the opened RTP reception port number is notified to the SIP server 11 ((6) in FIG. 9). Also, the SIP server 11 acquires the RTP reception port number received from the media server 14 and sets it as the RTP transmission destination from the media server.

  The SIP server 11 includes an IP address of the MRCP server used in the session (192.168.0.101 in the embodiment), an IP address of the media server used for transmission and reception of the MRCP channel ID and RTP (192.168.0.202 in the embodiment), a reception port The number is secured (step 94f), and this reception port number is set in the SDP (Session Description Protocol: RFC 4566) of the SIP “200 OK” message which is a response signal to the SIP “INVITE” message, and the voice application client (Step 94g and FIG. 9 (7)).

  FIG. 10 is a flowchart showing a control procedure of the telephone exchange device 30, the IVR 20, and the media server 14 when executing the speech synthesis process and the speech recognition process in the present embodiment.

  When performing the voice application process, the IVR 20 transmits an MRCP request to the MRCP server 12 (FIG. 10 (1)). When receiving the MRCP request, the MRCP server 12 determines from the content of the MRCP request that the voice dial service process is started (step 10a), and transmits a voice synthesis request to the media server 14 (FIG. 10 (2)). Then, a voice recognition request is transmitted to the media server 14 (FIG. 10 (3)). Upon receiving the request, the media server 14 gives a voice guidance transmission instruction to the voice synthesizer 1452 and prepares for voice guidance transmission (step 11a), and also gives a voice recognition preparation instruction to the voice recognition unit 1451 to recognize the sound. Preparation for processing is performed (step 11b).

  In the voice synthesizer 1452, for example, a voice guidance “Please input the name of the destination” is generated, and this voice guidance is transferred as an RTP packet to the IVR 20 via the channel of the session ID “1”. In FIG. 10, the RTP packet is converted into a voice signal and transferred to the caller telephone terminal 41 via the telephone exchange device 30 (FIG. 10 (4)). Therefore, the user of the telephone terminal 41 can know the next operation content by this voice guidance.

  Next, when receiving the voice signal, the IVR 20 transmits a SIP request for a voice recognition request to the SIP server 11 (FIG. 10 (5)). When the SIP server 11 recognizes that the request is a voice recognition request from the content of the SIP request, the SIP server 11 determines the MRCP server 12 by referring to the session management table of the session management DB 117 (step 12a), and sends the IVR 20 to the MRCP server 12. A voice recognition request is transmitted from (Fig. 10 (6)).

  When receiving the voice recognition request, the MRCP server 12 prepares to receive a voice recognition result (step 10b). After the reception preparation is completed, the MRCP server 12 notifies the IVR 20 that reception preparation is completed ((7) in FIG. 10).

  In this state, when the user of the telephone terminal 41 inputs a voice destination name “Yamamoto” in accordance with voice guidance, for example, a voice signal representing the name “Yamamoto” of the destination is sent from the telephone terminal 41 to the telephone. It is transmitted to the media server 14 via the exchange device 30 (FIG. 10 (8)).

  The media server 14 transfers the voice signal received from the telephone terminal 41 to the voice recognition unit 1451. The voice recognition unit 1451 performs voice recognition processing, creates a character string corresponding to the input voice signal using the voice-character string conversion table (step 11c), and is created as a result of the voice recognition process. The character string is notified to the MRCP server 12 ((9) in FIG. 10). The MRCP server 12 sets the notified “Yamamoto” character string in an MRCP response packet and transmits it to the IVR 20 as a recognition result of the voice signal (FIG. 10 (10)). If the MRCP server 12 cannot recognize the voice, the MRCP server 12 transmits a voice synthesis request to the media server 14 to generate a voice guidance saying “Could not be recognized.

  The IVR 20 extracts the character string “Yamamoto” from the MRCP response packet that is the voice recognition result transmitted from the MRCP server 12 and reads the telephone number “5621” corresponding to “Yamamoto”. Then, the control information composed of the telephone number and the transmission control signal is transferred to the telephone exchange device 30 as a text-based response (FIG. 10 (11)).

  The telephone exchange device 30 detects that the transmission control signal is included in the transferred control information and, based on the telephone number sent, performs, for example, dial transmission processing to the telephone terminal 41 of the transmission destination. Execute (step 13a). Thus, voice dialing is performed.

  In the above-described embodiment, the voice recognition preparation is performed before the voice guidance is transmitted from the media server 20, and the process that can cope with the voice input from the telephone terminal 41 during the voice guidance transmission is described. However, the present invention is not limited to the above-described operation. For example, voice guidance may be performed and voice recognition preparation may be performed after transmission.

  Here, during processing of request A from the telephone terminal 41 (shown in FIG. 1), as a request B shown in FIG. 1, for example, a function for executing a voice dialing function after the user of the telephone terminal 42 goes off-hook. The process when the number is input via the keypad or the function key will be described. The process described below is different only in the MRCP server and the media server assigned to the request B. The specific contents are as follows. Since it is the same as the process described in the request A from the telephone terminal 41 described in FIG. 9, the flowchart is omitted.

  First, a call request is notified to the telephone exchange device 30 together with a signal including the input function number.

  When the telephone exchange device 30 receives a call request from the telephone terminal 42, it performs incoming call transfer processing and transfers the call request to the IVR 20.

  The IVR 20 requests the SIP server 11 to generate a new session by SIP INVITE when a voice application process is required to respond to the request B, which is a request specified by the function number input from the telephone terminal 42.

  The SIP server 11 receives the request B for generating the SIP session at the communication unit 115, and searches the server in the “in operation” state from the MRCP server DB 113 managed by the server monitoring control unit 112 at the SIP control unit 11. If the MRCP server 13 is usable, the use of the MRCP server 13 is determined, and the session management DB 117 is associated with the session ID “2” as the session corresponding to the request B requested from the telephone terminal 42. The server ID “2” of the MRCP server 12 is registered.

  Then, the SIP server 11 notifies the MRCP server 12 of a session start via the interprocess communication control unit 116. Then, the MRCP server 12 acquires MRCP request information generated by the IVR 20 that is the voice application client notified from the SIP server 11 by the MRCP control unit 122 by starting a new session received via the inter-process communication control 123. Prepare to accept an MRCP request from the IVR 20.

  Subsequently, the SIP server 11 searches the media server DB 114 for servers that are in the “running” state, and if the media server 15 is usable, determines the use of the media server 15 by an arbitrary determination method. In the session management DB 117, the server ID “1” of the media server 15 is registered in association with the session ID “2” that is a session corresponding to the request B requested from the telephone terminal 42.

  Thereafter, the SIP server 11 notifies the media server 15 of the start of the session via the interprocess communication control 116.

  The media server 15 opens the RTP reception port from the IVR 20 that is the voice application client notified from the SIP server 11 by the RTP control unit 142 by starting a new session received via the inter-process communication control unit 143, The SIP server 11 is notified of the opened RTP reception port number. Further, the SIP server 11 sets the RTP reception port number received from the media server 15 as the RTP transmission destination from the media server.

  The SIP server 11 includes the IP address of the MRCP server used in the session (192.168.0.102 in the embodiment), the IP address of the media server used for transmission and reception of the MRCP channel ID and RTP (192.168.0.202 in the embodiment), the receiving port A number is secured, and this reception port number is set in the SDP (Session Description Protocol: RFC 4566) of the SIP “200 OK” message which is a response signal to the SIP “INVITE” message, and is sent to the voice application client IVR 20. Send.

  As described above, in the above embodiment, the voice application control unit 144 and the voice application engine unit 145 provided in the previous voice application server 10 are provided in another media server 14, and the MRCP control unit 122 is different from the media server 14. The SIP server 11 is distributed so as to have the SIP control unit 111 in the SIP server 11, and the SIP server 11 is connected to the usable media server 14 based on at least the voice application execution request from the telephone terminal 41. When a session for communication connection is established with the MRCP server 12, a session ID that identifies the session is associated with a server ID that identifies each of the media server 14 and the MRCP server 12 with which the session is established. Session The management table is stored and managed in the session management DB 117, and the session management table is referred to. Based on the reference result, a voice application such as voice synthesis service by voice synthesis or voice recognition is used for each request source for each session. The service will be executed.

  Therefore, by using the session management table, it is possible to specify a session to be used even when distributed as a device for each process, and when executing exchange service control involving voice application processing, the SIP server 11, MRCP server 12, and The media server 14 can cooperate and execute efficiently. Thereby, the load of the whole system can be reduced.

  In addition, since a session is fixedly prepared for each request source, even when requests for executing a plurality of voice applications from different telephone terminals 41 and 42 occur in the same time period, service control by these voice applications is performed. It can be executed efficiently in parallel.

  Further, according to the above embodiment, the SIP server 11 monitors the operation states of the media server 14 and the MRCP server 12, and the session is established only when the media server 14 and the MRCP server 12 are usable. It is possible to increase the effective utilization rate of the session by preventing unnecessary use.

  In addition, it is not limited to the above-described embodiment as it is, and in the implementation stage, the constituent elements can be modified and embodied without departing from the scope of the invention. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  DESCRIPTION OF SYMBOLS 10 ... Voice application system, 11 ... SIP server, 12, 13 ... MRCP server, 14, 15 ... Media server, 20 ... IVR, 30 ... Telephone exchange apparatus, 41-4n ... Telephone terminal, 111 ... SIP control part, 112 ... Server monitoring control unit 113 ... MRCP server DB, 114 ... media server DB, 116 ... inter-process communication control unit, 115 ... communication unit, 117 ... session management DB, 122 ... MRCP control unit, 142 ... RTP control unit, 144 ... Voice application control unit, 145 ... voice application engine part, INW ... IP network.

Claims (5)

A media server device that executes a predetermined voice application;
An application control server device connected to the media server device via an IP (Internet Protocol) network and controlling execution of the voice application;
Connected to the media server device and the application control server device via an IP (Internet Protocol) network and connectable to a telephone exchange device for performing call connection between a plurality of telephone terminals, the media server device and the application control A management server device that controls a session for communication connection between the server devices,
The management server device
When a session for communication connection is established between the media server device that can be used and the application control server device based on at least a voice application execution request from the telephone terminal, a session ID that identifies the session; Storage means for creating and storing a management table in which server IDs for specifying each of the media server device and the application control server device with which the session is established are associated with each other;
A voice application system comprising: processing means for performing processing using the voice application for a request source for each session based on the management table stored in the storage means. A telephone exchange for connecting a plurality of telephone terminals to a media server device that executes a predetermined voice application and an application control server device that controls execution of the voice application via an IP (Internet Protocol) network In a management server device capable of connecting a device and controlling a session for communication connection between the media server device and the application control server device,
When a session for communication connection is established between the media server device that can be used and the application control server device based on at least a voice application execution request from the telephone terminal, a session ID that identifies the session; Storage means for creating and storing a management table in which server IDs for specifying each of the media server device and the application control server device with which the session is established are associated with each other;
A management server apparatus comprising: processing means for performing processing using the voice application for a request source for each session based on the management table stored in the storage means.   The processing means transmits / receives a message defined by SIP (Session Initiation Protocol) with the request source for each session based on the management table when an audio signal is transmitted from the request source, 3. The management server device according to claim 2, wherein processing using a voice application is performed. Comprising monitoring means for monitoring operating states of the media server device and the application control server device;
The storage means is a communication connection between the usable media server device and the application control server device based on a monitoring result by the monitoring device when at least a voice application execution request is sent from the telephone terminal. A management table that associates a session ID that identifies the session with a server ID that identifies each of the media server device and the application control server device with which the session is established The management server device according to claim 2, wherein the management server device is created and stored. A telephone exchange for connecting a plurality of telephone terminals to a media server device that executes a predetermined voice application and an application control server device that controls execution of the voice application via an IP (Internet Protocol) network In a control method used in a management server device capable of connecting a device and controlling a session for communication connection between the media server device and the application control server device,
When a session for communication connection is established between the media server device that can be used and the application control server device based on at least a voice application execution request from the telephone terminal, a session ID that identifies the session; Storing a management table in which a server ID that identifies each of the media server device and the application control server device with which the session is established is associated;
A control method comprising: performing processing using the voice application for a request source for each session based on the stored management table.

Images