Dialogic Voice Cards - AutoCPA - Call Progress test tool

1 Issue overview
2 Recordings
3 Configuration
4 What are the problem types that AutoCPA can be used for?
5 Product list
6 GLOSSARY OF ACRONYMS / TERMS

Issue overview

The AutoCPA( ) tool allows the testing and tuning of various call progress parameters outside of a production environment by using a set of recordings that represent a real site. This type of information may be useful when troubleshooting call progress analysis issues, for example.

The AutoCPA tool performs call progress analysis (CPA) using the dx_dial API method with two voice resources (one voice resource for playing the file and the other for performing CPA) and recorded PCM files. The tool does not require any network resources and can be configured for various test scenarios. The tool can be loaded with multiple recorded PCM files and logs detection results directly to a command window and output text file for analysis.

Recordings

Users of the AutoCPA tool must gather recordings themselves for testing – the AutoCPA tool does not make the recordings itself. Recordings are normally gathered from another system in PCM 8K. The recordings should commence when the application starts call progress analysis and should end when the call has completed such that all the audio is captured.

Once completed, the PCM recordings can then be moved into the PCM directory of the AutoCPA tool (or where specified in the config.inf file – PCMdir field)

Configuration

The AutoCPA configuration (where applicable) is done through config.inf file, which is loaded upon initialization. Not all the parameters in the config.ini file are common between Dialogic DM3 series Media Boards and Dialogic JCT series Media Boards; for more information, visit the programming library documentation. The settings include:

General config: PCM directory, log output, voice resources.
DX_CAP Struct Settings
CPA Qualification Templates (JCT Media Boards only)
Global Tone Definitions (user and pre-defined)

The following settings are required to run the AutoCPA tool – all remaining settings will be set to their default values if commented out.

PCMdir = PCM                 # Directory where PCM recorded files are stored
LogFile = Results.txt     # Filename to log output
PlayDev = dxxxB1C1     # Voice device used for playing recorded PCM file
CallpDev = dxxxB1C2    # Voice device used for call progress

Usage and output:

Once the recordings have been gathered and the configuration file has been saved, run the AutoCPA.exe tool in a command prompt or double-click the file - no command line arguments are needed.

Once the tool has processed the PCM recordings in the specified PCM directory, the results will be written to the Results.txt file (or the location specified in the config.inf file – LogFile field). The results log file will contain the detected result for each recording in the following format:

[Recording directory] \ [recorded filename] * [detection result] * [detection time]

For example, the detection result: PCM\test.pcm * CON_PVD * 6.360 means: test.pcm was detected as a Positive Voice Detection (PVD) at 6.360 seconds into the file. The detection time can be useful, for example, to diagnose mis-detection issues that may have been caused by background noise.

What are the problem types that AutoCPA can be used for?

Mis-detections and tuning while using the DX method of CPA
DTMF mis-detections and tuning
Tone Definitions mis-detections and tuning

Product list

Dialogic JCT Media Boards

GLOSSARY OF ACRONYMS / TERMS

CPA – Call Progress Analysis
PVD – Positive Voice Detection
PAMD – Positive Answering Machine Detection