Voice Quality Monitoring via Alarm-Triggered Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Monitoring voice quality during teleconferences is challenging due to packet loss and delay variations in packet networks, which can lead to perceptual sound quality degradation, and existing methods may not accurately incorporate acoustic quality information, increasing computational burden, especially in mobile devices.

Innovation Solution

A method involving the analysis of transmission and acoustic quality metrics to determine voice quality, triggered by alarms exceeding threshold values, which includes sending diagnostic information and adjusting computational complexity based on alarm detection and estimation levels, allowing for improved accuracy and reduced computational load.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If comprehensive voice quality monitoring is implemented by analyzing transmission and acoustic quality metrics, then measurement precision is improved, but computational burden increases

Engineering Contradiction:
Improvevoice quality measurement accuracyVSAvoidcomputational burden
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The voice quality monitoring system is segmented into multiple independent modules: transmission quality analysis module that processes packet metadata (jitter, packet loss), acoustic quality analysis module that processes audio signals (echo, noise), and a lightweight integrator that combines results. This segmentation allows each module to operate independently with optimized computational requirements, reducing overall system complexity while maintaining comprehensive monitoring capability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system implements selective monitoring by analyzing only the necessary components of voice quality based on current network and acoustic conditions. Rather than continuously processing all audio signals at full computational intensity, the system adjusts the depth of analysis based on alarm thresholds and current quality levels, performing comprehensive analysis only when needed while using lighter monitoring during stable conditions.

Inventive Principle:
Principle #16Partial or excessive action

2Reliability

If real-time voice quality monitoring is implemented during teleconferences, then reliability is improved, but loss of time increases due to processing delays

Engineering Contradiction:
Improvevoice quality monitoring reliabilityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary analysis of transmission quality metrics (jitter, packet loss rates) and acoustic environment parameters continuously in the background before actual voice quality degradation occurs. By pre-processing and pre-analyzing these parameters, the system prepares quality assessments in advance, enabling rapid response when actual degradation is detected without requiring intensive real-time processing during critical moments.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements adaptive sampling rates and processing intervals that skip detailed analysis during stable quality periods and rush through comprehensive analysis only when quality thresholds are breached. This allows the system to maintain reliability by focusing computational resources on critical moments when voice quality actually degrades, rather than uniformly processing all time periods at full intensity.

Inventive Principle:
Principle #21Skipping (Rushing through)

Data Source

PatentUS9917945B2In-service monitoring of voice quality in teleconferencing
Publication Date: 2018.03.13 DOLBY LABORATORIES LICENSING CORP
  • US9917945B2 patent drawing
  • US9917945B2 patent drawing
  • US9917945B2 patent drawing

AI summary

A voice quality estimation process may be triggered by receiving one or more alarms corresponding to one or more endpoint terminals being used during a teleconference. The alarm(s) may include uplink transmission alarms, downlink transmission alarms and/or acoustic quality alarms. The alarms may be based on evaluating transmission metrics and/or acoustic quality metrics. The voice quality estimation process may require a relatively greater computational burden than the processes of evaluating the transmission metrics and/or acoustic quality metrics for the purpose of potentially triggering an alarm. The accuracy and computational complexity of voice quality estimation may be adjusted by selecting times during which alarm detection will take place, alarm detector thresholds, alarm analyzer thresholds and/or levels of voice quality estimation.