You are here

Voice Quality and Call Completion Rate for an Operational Radio Test


TECHNOLOGY AREA(S): Info Systems, 

OBJECTIVE: Provide comprehensive call completion rate determination during an operational radio test by obtaining objective measures of the voice quality of each call made. This requires instrumenting radios/operators and automatically grading the voice quality and call completion rate. 

DESCRIPTION: This effort requires developing instrumentation to collect the voice traffic; harvest, manage, and correlate the voice data; and automatically determine the quality of the received voice traffic to determine call completion rate. The requirement can be subdivided into the following four areas: • Voice Collection: The ability to capture all outgoing and incoming voice calls with minimum to no interference with the radio operator. This requires any appended collection capability to have a very small size and weight while having enough power and data storage to not require battery changes or data harvesting within a 96-hour test window. This collection must capture the voice in a manner that is representative of what the Soldier hears. Collecting voice-over IP (VOIP) traffic will not meet this requirement because it does not address the quality of the system under test microphone or speakers which are key components of effective voice quality. • Voice Data Harvesting and Management: The ability to harvest and manage all collected voice as well as capture key metadata about the voice data (e.g. time collected, location, unit, radio identification, etc.). • Voice Data Correlation: The ability to correlate all outgoing/sent voice calls with their corresponding received calls. A call is considered a single transmission sent/received pair. The metadata collected must be sufficient to perform this correlation of the voice data files. • Voice Scoring: The ability to automatically apply a scoring algorithm to compare the sent with received voice data. An example of a scoring algorithm is the OPTICOM Perceptual Objective Listening Quality Analysis algorithm. The voice scoring algorithm must be adjustable to allow for calibration to a user-determined acceptable level. This would be done by conducting a manual voice quality test and comparing the manual scoring to the automated scoring algorithm result. From this comparison, a pass or fail value would be determined and the voice scoring algorithm would be calibrated to that user-determined acceptable level. Finally, the system will need to generate reports showing call origination and reception over time and call-completion rates over time based on the pass/fail value provided. 

PHASE I: Develop a concept for a prototype voice collection capability and present how it will be capable of meeting size, weight, and power constraints in an operational environment. Develop a design for voice data harvesting, management, and correlation to allow for voice scoring. Demonstrate a voice scoring algorithm using government-supplied sent/received voice traffic. 

PHASE II: Develop and demonstrate a prototype voice collection capability. Develop a prototype voice data harvesting, management, and correlation capability to allow for voice scoring. Implement a prototype voice scoring algorithm and reporting capability. 

PHASE III: The end-state of this effort will be to transition and mature this research and prototype capability to deployable test instrumentation for evaluating voice communications during operational testing of new tactical radio systems (e.g. Leader Radio, Manpack). Commercial applications: Conducting voice assessment of customer service call center agents to verify they can be easily understood. Automated assessment of students learning to speak a new language. 


1: Beerends, John & Schmidmer, Christian & Berger, Jens & Obermann, Matthias & Ullmann, Raphael & Pomy, Joachim & Keyhl, Michael. (2013). Perceptual Objective Listening Quality Assessment (POLQA), The Third Generation ITU-T Standard for End-to-End Speech Quality Measurement Part I-Temporal Alignment. AES: Journal of the Audio Engineering Society. 61. 366-384.

KEYWORDS: Voice Collection, Voice Scoring, Call Completion Rate, Radio, Operational Test, Voice Correlation 

US Flag An Official Website of the United States Government