Call Quality Monitoring With ML-Weighted MOS, PDD, And Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing call quality measurement techniques provide limited visibility and accuracy, as they often rely on single-point monitoring or active testing, which may not reflect actual user experiences and can be influenced by network dynamics and device variations.
Innovation Solution
A system and method for monitoring call quality using a composite quality score based on Mean Opinion Score (MOS), Post-Dial Delay (PDD), and latency, calculated at a single point in the communication path, incorporating machine learning for parameter weighting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If active testing is used to measure call quality, then measurement accuracy for the tested communication path is improved, but the ability to reflect actual user experiences deteriorates
Solution Approach 1:
The patent creates a virtual copy of the actual call path by routing test calls through the same network elements and devices as real user calls. This virtual replica allows controlled measurement while maintaining representativeness, as the test infrastructure mirrors the actual production environment including network topology, device types, and routing paths.
Solution Approach 2:
The patent segments the call quality measurement into multiple independent measurement points along the communication path. By placing measurement nodes at different locations (e.g., at network elements, at endpoints), the system can measure specific segments of the call path independently, allowing comprehensive coverage of the entire call flow while maintaining control over each measurement point.
2Quantity of substance
If all calls are monitored at a single point, then measurement coverage is improved, but visibility into the complete call path deteriorates
Solution Approach 1:
The patent divides the call path into multiple segments by placing measurement points at different network elements and endpoints. Each measurement point captures information for its specific segment, and the system aggregates these segmented measurements to reconstruct the complete call path picture, maintaining both comprehensive coverage and full visibility.
Solution Approach 2:
The patent adds a temporal dimension to single-point monitoring by measuring calls at multiple time points throughout the call duration. The system captures measurements at call setup, during active conversation, and at call termination, providing a multi-dimensional view of call quality evolution while maintaining focus at each measurement point.
3Loss of information
If call quality is measured at multiple points in the path, then complete call path visibility is improved, but system complexity and resource requirements deteriorate
Solution Approach 1:
The patent implements a universal measurement framework where the same measurement infrastructure and processing logic can be deployed at multiple points in the call path. The measurement points use standardized protocols and the same analysis algorithms, allowing multi-point monitoring without proportionally increasing system complexity, as each point operates as an independent instance of the same system.
4Speed
If latency is reduced for real-time monitoring, then responsiveness is improved, but measurement accuracy under varying network conditions deteriorates
Solution Approach 1:
The patent performs preliminary measurements and calculations during call setup and initial phases, establishing baseline metrics before the call reaches critical quality assessment points. This allows the system to have ready-to-compare reference values when actual quality degradation occurs, enabling rapid detection without waiting for complete call analysis.
Data Source
AI summary
A system for monitoring call quality may include a monitoring environment. The monitoring environment may be operable to obtain information associated with a call between two or more computing devices, and may generate a composite quality score for the call based on the information. The composite quality score may be based on values of a plurality of parameters associated with the call. A method of monitoring call quality is also disclosed.