Jump to: navigation, search

Configuring the GVP Components

Perform these advanced configuration procedures after GVP installation and basic configuration.

Integrating Application Objects

After the Media Control Platform and Call Control Platform Application objects are created and the components are installed, they are integrated with the Resource Manager which acts as a proxy server. SIP devices and VoiceXML or CCXML applications can then make use of media-centric services through the proxy, without having to know the actual location of these resources.

This procedure is optional and required only if you want the Resource Manager to act as a proxy server for outbound requests. To integrate these Application objects with the Resource Manager, you configure the Session Initiation Protocol (SIP) settings.

This procedure describes how to integrate Application objects with the Resource Manager by configuring SIP and secure SIP options.

Procedure: Integrating Application Objects with Resource Manager

Integrate an Application object with Resource Manager by configuring the Application parameters.

  1. Verify that all GVP components are installed. See Procedure: Using the Deployment Wizard to Install GVP.
  2. Log in to Genesys Administrator.
  3. On the Provisioning tab, select Environment > Applications.
  4. Click the Application object that you want to configure for example, the Media Control Platform or Call Control Platform Application.
    The Configuration tab appears.
  5. Click the Options tab, and use the View drop-down list to select Show options in groups...
  6. In the sip section, find the routeset option.
  7. In the Value field of the routeset option, enter the following:
    • sip:<IP_RM>:<SIPPort_RM>;lr
      to integrate the Media Control Platform with Resource Manager.
    • sip:<IP_RM>:<SIPPort_RM>;lr
      to integrate the Call Control Platform with Resource Manager.

      ...where IP_RM is the IP address of the Resource Manager, and SIPPort_RM is the SIP port of the Resource Manager typically, 5060.

      For both MCP & CCP the syntax is the same:
      routeset = sip:<RM1_IP>:<RM1_proxy_port>;lr
      for half of the MCPs & CCPs, the other half use
      routeset = sip:<tt><RM2_IP>:<RM2_proxy_port>;lr</tt>

      (Where anything not in italics is required verbatim.)
      ...where proxy_port is the port from [proxy] sip.transport.n


      Tip
      You must include the angle brackets in the Value field in the sip.routeset and sip.securerouteset parameters.
    • In the Value field of the securerouteset option, enter the following:
      • sip:<IP_RM>:<SIPSecurePort_RM>;lr to integrate the Media Control Platform with Resources Manager.
      • sip:<IP_RM>:<SIPSecurePort_RM> to integrate the Call Control Platform with Resource Manager.
        ...where IP_RM is the IP address of the Resource Manager, and SIPSecurePort_RM is the SIP secure port of the Resource Manager typically, 5061.
        Tip
        The GVP components and the external SIP Server support secure SIP capabilities.
    • To use the Call Recording Solution through third-party recording servers: In the vrmrecorder section, configure the following options (pointing to the Resource Manager's IP address and SIP port, as shown in the two previous steps:
      • sip.routeset
      • sip.securerouteset
    • Save the configuration.
    • Create the connections to the Message Server. See Procedure: Creating a Connection to a Server.

Creating a Connection to a Server

Use the procedure in this section to create connections to:

  • The Message Server—Create a connection in the Media Control Platform, Call Control Platform, Resource Manager, Supplementary Services Gateway, CTI Connector, PSTN Connector, MRCP Proxy, Reporting Server and Policy Server Applications to ensure that component log information reaches the Log database and can be viewed in the Solution Control Interface (SCI).
  • The Reporting Server—Create a connection in the Media Control Platform, Call Control Platform, Resource Manager, PSTN Connector, CTI Connector, Supplementary Services Gateway, and MRCP Proxy Applications to ensure that these components detect the Reporting Server to which they are sending reporting data. Genesys Administrator also requires a connection to Reporting Server to monitor GVP components.
  • SIP Server—Create a connection in the Resource Manager, Supplementary Services Gateway, and PSTN Connector Applications to manage the initiation of outbound calls.
  • MRCP Proxy—Create a connection in the Media Control Platform Application if you are planning to use the proxy to manage MRCPv1 RTSP traffic within the GVP deployment.
  • MRCP Server—Create a connection in the MRCP Proxy Application if you are planning to use the proxy to manage MRCPv1 RTSP traffic within the GVP deployment (in the Media Control Platform Application if you are not deploying the MRCP Proxy).
  • Cisco T-Server—Create a connection in the UCM Connector Application to ensure the tenant DBID of the Cisco T-Server is included in Request URI in any SIP INVITE messages sent to the UCM Connector.
  • The SNMP Master Agent—Create a connection in the Media Control Platform, Call Control Platform, Resource Manager, Supplementary Services Gateway, CTI Connector, PSTN Connector, MRCP Proxy, and Reporting Server Applications if you want to capture alarm and trap information.

Procedure: Creating a Connection to a Server

This procedures creates a connection in an Application object to a server or component.

  1. Verify that all GVP components are installed. See Procedure: Using the Deployment Wizard to Install GVP
  2. Log in to Genesys Administrator.
  3. On the Provisioning tab, select Environment > Applications.
  4. Click the Application object for which you are creating the connection; for example, the Media Control Platform Application object.
    The Configuration tab appears.
  5. In the General section, in the Connections field, click Add.
    The Connection Info dialog box appears.
  6. In the Server field, click the down arrow to open the Browse Application dialog box.
  7. Select the server or component to which you want to create a connection for example, Message Server, SIP Server, or SNMP Master Agent. The required fields in the Connection Info section are populated automatically. (Ensure the Connection Protocols field is left blank. It is not required for GVP components.)
  8. Click OK.
    The server or component you selected in Step 6 appears under Connections.
  9. Save the configuration.
  10. Complete the remaining post-installation activities for the Media Control Platform. See Procedure: Provisioning Speech Resource Application Objects.

Provisioning the Speech Resources

The Media Resource Control Protocol (MRCP) speech resources are controlled by the Call Manager Application Program Interface (CMAPI), which opens and closes sessions, and provides the speech recognition and speech synthesis commands that the MRCP Server uses to carry out speech requests.

If the MRCP Proxy is deployed, the configurations in this procedure vary slightly. Therefore, the configurations are described with and without the MRCP Proxy. If you have installed the MRCP Proxy, see also Provisioning the MRCP Proxy.

Tip
The procedures in this section are required only if you are using Automatic Speech Recognition (ASR) and Text-to-Speech (TTS) speech resources, and have an MRCP Server or MRCP Proxy in your deployment.

This section contains two procedures that create the Speech Resource Applications and assign the MRCP Server or MRCP Proxy to the Media Control Platform.

Procedure: Provisioning Speech Resource Application Objects

This procedure creates the MRCP Speech Resource Applications for ASR and TTS. After a Speech Resource Application is created with the basic configuration, it must be provisioned with the IP address and port number of the MRCP Server or the MRCP Proxy (if required).

  1. Verify that:
  2. Log in to Genesys Administrator.
  3. On the Provisioning tab, select Environment > Applications.
  4. Select the Speech Resource Application you want to configure.
    The Configuration tab appears.
  5. Click the Options tab, and scroll to the provision section.
  6. Enter the value for each Option as described in this table:
    Table: MRCP Application Properties Options Tab
    Option name Option value
    For MRCPv1
    vrm.client.resource.name Enter the identifier used to link the VoiceXML application to a common set of speech resources; for example,
    • For ASR, enter ATT, IBM, LOQUENDO, LUMENVOX, NUANCE, TELISMA, or ONMOBILE.
    • For TTS, enter ATT, ACAPELA, VOCALIZER, IBM, or VOXYGEN.

    Notes:

    • A common set of speech resources means that the provisioning data for each speech resource with the same name is identical. A resource with the same name but different provisioning data should not be added to the common set of resources.
    • GVP supports dynamically removing and adding ASR/TTS servers (resources), but does not support dynamically changing a resource's provisioning data. To change a resource's provisioning data, follow these steps:
      1. Remove (delete) the resource.
      2. Modify the resource s provisioning data.
      3. Add back (reconnect) the resource.
      Important: If the provisioning data of the modified resource is different from an existing common set of resources with the same name, then you must use a different name for this resource.
    vrm.client.resource.uri The URI must contain the IP address and port number of the MRCP Server by using the following format:

    rtsp://servername:<port>/<path> For the recommended resource Uniform Resource Identifier (URI), check the MRCP vendor documentation.

    Tip
    The MRCP Proxy supports MRCPv1 speech resources only.
    vrm.proxy.ping_interval Enter a value (or retain the default) to specify the ping interval in milliseconds (used only when the MRCP Proxy is deployed).

    Default value: 30000

    For MRCPv2
    vrm.client.resource.name Enter the identifier used to link the VoiceXML application to a common set of speech resources for example,

    For ASR, enter NUANCE or RECOGNIZER.
    For TTS, enter NUANCE or VOCALIZER.

    vrm.client.resource.uri The URI must contain the IP address and port number of the MRCP server using one of two formats:

    sip:mresources@<MRCP server IP>:<port>;transport=TLS
    sips:mresources@<MRCP server IP>:<port>
    (The default SIPS port number for Nuance Speech Servers is 5061.)
    For the recommended resource URI, check the MRCP vendor documentation.

    vrm.client.TransportProtocol Enter one of two values:

    MRCPv2 without Security
    MRCPv2 with secure TLS

  7. Save the configuration.
    Tip
    Complete Steps 8 to 11 if you are deploying MRCPv2 with Secure RTP (SRTP) only.

    Configure the Media Control Platform Application

  8. Select the Media Control Platform Application that is associated with this speech resource.
    The Configuration tab appears.
  9. Click the Options tab, and scroll to the mpc section.
  10. Configure the following parameters with the values that are shown here:
    • asr.srtp.mode=offer
    • asr.srtp.sessionparams=none
    • tts.srtp.mode=offer
    • tts.srtp.sessionparams=none
  11. Save the configuration.
    Configure the ASR Server
  12. Configure the following options on the ASR Server:
    • If the ASR Server supports session timeout, configure 600000 (milliseconds, or 10 minutes) for the timeout value to prevent interruption of any active recognition sessions.
    • For Nuance SpeechWorks MediaServer and OpenSpeech Recognizer, configure the server.transport.sessionTimeout VXIInteger option with a value of 600000 (10 minutes).
    • For Nuance Speech Server and Nuance Recognizer, configure the server.mrcp2.sip.sessionTimeout and server.mrcp1.rtsp.sessionTimeout options with a value of 600000 (10 minutes).
      For other ASR vendors, check the vendor documentation.
  13. To make the ASR service work correctly with GVP, you must edit the Nuance Recognizer file baseline.xml. and comment out the fourth and fifth lines in the code sample below:
    <param name="swirec_extra_nbest_keys">
    <declaration group="result" type="string_set" set_by="default+api"> </declaration>
    <value>SWI_meaning</value>
    <value>SWI_literal</value>
    <value>SWI_grammarName</value>
    </param>

    The characters to add to the code are marked in red.
  14. Assign the MRCP Server to the Media Control Platform Application object. See Procedure: Assigning the MRCP Server to the Media Control Platform.

Procedure: Assigning the MRCP Server to the Media Control Platform

Use this procedure if you have not deployed the MRCP Proxy, otherwise see Provisioning the MRCP Proxy.

  1. Verify that:
  2. Log in to Genesys Administrator.
  3. On the Provisioning tab, select Environment > Applications.
  4. Double-click the Media Control Platform Application object that you want to configure.
    The Configuration tab appears.
  5. In the General section, in the Connections field, click Add.
    The Connection Info dialog box appears.
  6. Enter the information in the required fields, as shown in this table:
    Table: Connection Info Dialog Box
    Field Description
    Server Browse to select the MRCP Server.
    ID This field is populated automatically with the value default.
    Trace Mode This field is populated automatically with the value Trace is Turned Off.
    Application Parameters Enter provisiontype=primary for a primary MRCP server.

    Enter provisiontype=backup for a backup MRCP server.

  7. Click OK.
  8. Save the configuration.
    Tip
    There is no limit to the number of primary or backup MRCP servers that you can assign to the Media Control Platform; however, do not assign the same server as both primary and backup.
  9. If required, complete the post-installation activities for the Supplementary Services Gateway. See Provisioning the PSTN Connector.

Enabling Native Connection to Google Cloud Speech-to-Text in MCP

Prerequisites for using the feature

  • You must open a Google Cloud Speech-to-Text account with Google.
  • Get a Google Speech-to-Text Credential file from Google and place it anywhere on the MCP host. This is the same credential file as used for Text-to-Speech (TTS). Ensure the user who runs MCP has the READ permission.

Configuration/provisioning steps

Important
  1. Step 1-5 can be skipped if Google Speech-to-Text has been enabled in MCP.
  2. You must restart MCP whenever the Google credential file is updated, or the option [vrm]client.grpc.credential is changed.

To enable the Native Connection feature to Google Cloud Speech-to-Text in MCP:

  1. Import the VP_MCP_GRPC_ASR_900 template from the MCP IP package.
  2. Create a Resource Access Point application from the VP_MCP_GRPC_ASR_900 template. If you plan to use a proxy, set the [provision]vrm.client.resource.proxy option to the proxy in the following format http://host:port. Use the default settings for other provision parameters.
  3. Add a connection to the Resource Access Point application in MCP.
  4. Get a Google Speech-to-Text credential file from Google and place it anywhere on the MCP host. This is the same credential file as used for TTS. Ensure the user who runs MCP has the READ permission.
  5. Set the MPC option [vrm]client.grpc.credential to the path of the Google Speech-to-Text Credential file, for example, C:\Program Files\GCTI\gvp\VP Media Control Platform 9.0\MCP900\config\google.json.
  6. Add GRPC into the MCP option [vrm]client.modules.
    Important
    MCP should change [asr] load_once_per_call=0 to support GSR.
  7. Restart MCP.

To enable Google Speech-to-Text for VoiceXML applications, apply the following changes to the GASR Resource Access Point:

  1. Set vrm.client.resource.name provision parameter to nativegsr.
  2. Set vrm.client.grammar.expected provision parameter to builtin:grammar/transcribe.

In VoiceXML, Google Speech-to-Text can be triggered by setting the com.genesyslab.asrengine VoiceXML property and src attribute in the <grammar> tag to match the values provisioned in the GRPC Resource Access Point. For example:

<?xml version="1.0"?>
<vxml version="2.0" xmlns="http://www.w3.org/2001/vxml">
  <form id="Welcome">
    <property name="com.genesyslab.asrengine" value="nativegsr"/>
    <block>
        <prompt> What do you want to say? </prompt>
    </block>
    <field name="pword">
        <grammar mode="voice" xml:lang="en-US" src="builtin:grammar/transcribe" />
        <filled>
          <prompt>
           You said <value expr="pword" />.
          </prompt>
          <exit/>
        </filled>
    </field>
  </form>
</vxml>

Google Speech-to-Text in VoiceXML applications does not support phase hints, except in a scenario such as speech_contexts[] vendor-specific property. For example, <property name=”com.genesyslab.asr.speech_context” value=”apple, orange”/>

To improve recognition results, Google Speech-to-Text in VoiceXML applications supports N-best. It does not support complete and incomplete timeout.

VoiceXML application limitations with GSR

Google Speech Recognition (GSR) doesn’t follow the ASR specifications in the VoiceXML world. It cannot fit in VoiceXML applications seamlessly like MRCPv1/2 ASR. Thus, there are a few limitations that include the following:

  1. No support for speech grammars ( Speech Recognition Grammar). Any grammar specified in VoiceXML application will be ignored when GSR is used. However, there must be at least a grammar (it can be a duplicate one) specified in VoiceXML application because NGi requires it to launch an ASR session.
  2. No support for input slot because of 1.
  3. No support for recording utterance because GSR cannot provide recordings.
  4. No support for Offboard DTMF recognition.
  5. Support only for load_once_per_call=0 (false) due to the nature of GSR.
  6. It would be expected that GSR recognition result may be unexpected due to the lack of grammar support.
  7. Due to 6, no support for <menu> to be used in VoiceXML application that is served by GSR.

Enabling Native Connection to Google Cloud Text-to-Speech in MCP

Prerequisites for using the feature

  • You must open a Google Cloud Text-to-Speech account with Google.
  • Get a Google Text-to-Speech credential file from Google and place it anywhere on the MCP host. This is the same credential file as used for ASR. Ensure the user who runs MCP has the READ permission.

Configuration/provisioning steps

Important
  1. Step 1-5 can be skipped if Google Speech-to-Text has been enabled in MCP.
  2. You must restart MCP whenever the Google credential file is updated, or the option [vrm]client.grpc.credential is changed.

To enable the Native Connection feature to Google Cloud Text-to-Speech in MCP:

  1. Import the VP_MediaControlPlatform_900 template from MCP IP package.
  2. Create a MCP application and install MCP.
  3. Get the credential file of your Google account from Google and place it anywhere on the MCP host. The credential is shared with ASR. Ensure the user who runs MCP has the READ permission.
  4. Set the MPC option [vrm]client.grpc.credential to the path of the Google credential file, for example, C:\Program Files\GCTI\gvp\VP Media Control Platform 9.0\MCP900\config\google.json.
  5. Add GRPC into the MCP option [vrm]client.modules.
  6. Import the VP_MCP_GRPC_TTS_900 template from the MCP IP package.
  7. Create a Resource Access Point application from the VP_MCP_GRPC_TTS_900 template. If you plan to use a proxy, set the [provision]vrm.client.resource.proxy option to the proxy in the following format http://host:port. Use the default settings for other provision parameters.
  8. Add a connection to the Resource Access Point application in MCP.
  9. Restart MCP.
Important
To engage Google TTS, you need to explicitly specify the TTS engine, either through the com.genesyslab.ttsengine property or through MCP [tts] defaultengine.

The default resource name of Google Text-to-Speech is “GTTS”. It is set in the Resource Access Point application created in the above step 7.

Sample VXML for validation

<?xml version="1.0"?>
<vxml version="2.1" xmlns="http://www.w3.org/2001/vxml">
<property name="com.genesyslab.ttsengine" value="GTTS" />
<form id="Welcome">
<block name="Hello">
<prompt>
Hello world!
</prompt>
</block>
</form>
</vxml>

Troubleshooting tips for common issues

Specific properties applied to Google Text-to-Speech

  1. <property name="com.genesyslab.tts.speaking_rate" value="1.0"/>
    • Optional speaking rate/speed, in the range [0.25, 4.0].
    • 1.0 is the normal native speed supported by the specific voice.
    • 2.0 is twice as fast.
    • 0.5 is half as fast.
  2. <property name="com.genesyslab.tts.pitch" value="0.0"/>
    • Optional speaking pitch, in the range [-20.0, 20.0].
    • 20 means increase 20 semitones from the original pitch.
    • -20 means decrease 20 semitones from the original pitch.
  3. <property name="com.genesyslab.tts.volume_gain_db" value="0.0"/>
    • Optional volume gain (in dB) of the normal native volume supported by the specific voice, in the range [-96.0, 16.0].
    • If unset, or set to a value of 0.0 (dB), will play at normal native signal amplitude.
    • A value of -6.0 (dB) will play at approximately half the amplitude of the normal native signal amplitude.
    • A value of +6.0 (dB) will play at approximately twice the amplitude of the normal native signal amplitude.
    • Genesys strongly recommends not to exceed +10 (dB) as there's usually no effective increase in loudness for any value greater than that.
  4. <property name="com.genesyslab.tts.name" value="en-US-Wavenet-A"/>
    • The name of the voice. Optional.
    • If not set, the service will choose a voice based on the other parameters such as language (set in the xml:lang attribute of <prompt>) and gender (set in the gender attribute of <voice>). The voice list can be found here.

Google Text-to-Speech unchangeable options

  1. PCM16 encoding will be used between MCP and Google Text-to-Speech.
  2. 8000Hz sample rate will be used.
  3. "telephony-class-application" will be used as the audio effect profile. See this page for information on telephony-class-application.

Provisioning the MRCP Proxy

The MRCP Proxy is an optional component, but must be deployed if ASR and TTS usage reporting is required. You can deploy the MRCP Proxy in stand-alone or warm active standby HA mode. The procedures in this section describe the steps for each configuration.

Tip
By design, the MRCP Proxy supports only the NUANCE speech resource.

Procedure: Configuring the MRCP Proxy

Configure the MRCP Proxy to act as a proxy for all MRCPv1 traffic in the environment.

  1. Verify that:
  2. Log in to Genesys Administrator.
  3. On the Provisioning tab, select Environment > Applications.
  4. Double-click the MRCP Proxy Application that you want to configure.
    The Configuration tab appears.
  5. Click the Options tab, in the vrmproxy section, configure the host part of the uri configuration option with the actual IP address of the MRCP Proxy.
    Important
    If the Media Control Platform is installed on the same host as the MRCP Proxy, retain the default value for the uri configuration option.
  6. Create a connection to the MRCP Server. See the Prerequisites section of this procedure.
  7. Save the configuration.

Procedure: Configuring the MRCP Proxy for HA

A configured MRCP Proxy acts as a warm standby in case of failover which means that, like a hot standby, the standby instance becomes active if the active instance fails. However, unlike a hot standby, a warm standby does not handle existing sessions. Application requests are rejected mid-stream during a failover; and applications must be designed to accommodate such a failure.

The failover sequence of events is as follows:

  1. The primary MRCP Proxy terminates.
  2. The LCA in the primary MRCPP machine informs SCS about this event.
  3. SCS checks to see if the terminated MRCPP has a backup instance configured.
  4. If there is a backup instance configured, SCS instructs through LCA in the backup computer the other MRCPP to become primary.

In a standard configuration, the MRCP Proxies are configured as backup to each other, and SCS has an HA license to perform a switch-over.

This procedure configures the MRCP Proxy in HA mode to act as a proxy for all MRCPv1 traffic in the environment.

  1. Verify that:
  2. Complete Procedure: Configuring the MRCP Proxy for both MRCP Proxy Applications in the HA pair.
  3. In the primary MRCP Proxy Application, click the Configurations tab.
  4. In the Server Info section, in the Backup Server field, browse to the backup MRCP Proxy Application and click to select it.
  5. In the Redundancy field, select warm-standby.
  6. Save the configuration.
    Connect to the MCP
  7. In the Media Control Platform Application, create a connection to the primary MRCP Proxy.
  8. Save the configuration.

Procedure: Adding a Speech Server as Primary or Backup

  1. Log in to Genesys Administrator.
  2. On the Provisioning tab, select Environment > Applications.
  3. Select the MRCP Proxy Application that you want to configure and click Manage Connections.
    The Manage Connections dialog appears.
  4. Click Next twice in the Manage Connections dialog.
    The Add Connections dialog appears.
  5. Click Add and select the speech server to add.
  6. Click Edit and select the Advanced tab.
  7. Enter provisiontype=primary in the Application Parameters field, to add the speech resource as primary.
    OR
    Enter provisiontype=backup in the Application Parameters field to add the speech resource as backup.
  8. Click Execute and then Finish.
Retrieved from "https://docs.genesys.com/Documentation:GVP:GDG:CGCGC:9.0.x (2019-08-23 07:01:11)"
This page was last modified on August 3, 2019, at 03:28.

Feedback

Comment on this article:

blog comments powered by Disqus