Bi-directional Audio Generation Device for Telephony Quality Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for assessing speech quality in telephony systems require live user involvement, making them time-consuming and expensive, and are limited to dedicated tests in live networks.
Innovation Solution
An automated electronic device connected to a telephony network component that uses a bi-directional audio generation system to perform Perceptual Evaluation of Speech Quality assessments without live user involvement, calculating Mean Opinion Scores using a Full Reference algorithm and storing results for remote access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated electronic device is used for speech quality assessment, then productivity and cost-efficiency are improved, but measurement precision may be compromised compared to human listeners
Solution Approach 1:
The patent uses Full Reference (FR) algorithms that create a digital copy of the reference speech signal and process it through the same codec under test, enabling automated comparison without human involvement. This copying approach allows the automated device to achieve measurement precision comparable to human listeners by mathematically analyzing the differences between original and decoded signals.
Solution Approach 2:
The patent replaces the mechanical system of human listeners physically gathering and subjectively scoring speech quality with an automated electronic system that uses digital signal processing algorithms (PESQ, POLQA) to objectively measure and calculate Mean Opinion Scores. This substitution maintains measurement precision while dramatically improving productivity and eliminating the need for human involvement.
2Measurement precision
If Full Reference algorithm is used, then measurement precision is improved, but device complexity increases due to need for reference signal and synchronized processing
Solution Approach 1:
The patent merges the reference signal generation, encoding, decoding, and comparison functions into a single automated electronic device. By combining these functions internally, the device eliminates the need for separate reference signal sources and synchronized processing systems, thereby reducing device complexity while maintaining Full Reference algorithm precision.
3Loss of time
If automated testing is implemented, then loss of time is reduced, but adaptability decreases due to limitation to dedicated tests in live networks
Solution Approach 1:
The patent designs the automated electronic device with multi-functionality to perform various speech quality assessment tasks including different codec evaluations, network condition testing, and algorithm comparisons. This universal design allows the device to adapt to different testing scenarios and requirements while maintaining automated efficiency, thereby improving both adaptability and reducing time loss simultaneously.
Data Source
AI summary
A system for bi-directional quality testing of a telephony system or network using an audio generation device configured to automate Perceptual Evaluation of Speech Quality assessments and connect to a communication endpoint through computer telephony integration (CTI) instruction or system configuration to enable auto-answering of the communication endpoint, listen, record, and process signals and audio data, calculate Mean Opinion Scores (MOS), generate signals and audio for playback, compare files using a full reference algorithm, and store data with MOS results displayed in data naming structure.