Voice Activity-Aware Audio Quality Evaluation for VoIP Calls
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating audio quality in voice-over-internet-protocol (VoIP) services in telecommunication networks are inadequate, as they rely solely on packet loss rate and muting rate, failing to account for voice activity levels, which can lead to inaccurate user perception assessments.
Innovation Solution
Implementing network components like UPF in 5G core networks and SGW/PGW in 4G EPC networks to measure packet loss rate, muting rate, and voice activity levels, using RTP packet analysis to determine user perception impact, and generate alerts for potential quality issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If packet loss rate and muting rate are used to measure audio quality, then the measurement is simple, but the accuracy of user perception assessment is insufficient
Solution Approach 1:
The patent introduces voice activity detection as an intermediary metric that mediates between raw packet loss data and user perception assessment. By detecting whether packets contain actual voice content or silence, the system creates an intermediate layer of analysis that improves accuracy without requiring direct complex user feedback mechanisms
Solution Approach 2:
The patent transforms the audio quality assessment by changing the parameter being measured from simple packet loss count to voice activity-aware packet loss. This parameter change allows the same measurement infrastructure to provide more accurate results by considering the content type (voice vs. silence) of lost packets
2Measurement precision
If voice activity detection is added to audio quality evaluation, then the accuracy of user perception assessment is improved, but the complexity of the evaluation system increases
Solution Approach 1:
The patent makes the existing network infrastructure multi-functional by enabling it to perform both traditional packet loss monitoring and voice activity detection using the same packet inspection mechanisms. This universality allows improved measurement accuracy without proportionally increasing system complexity
Solution Approach 2:
The system performs self-service by using the audio packets themselves to determine voice activity status. The same packets that carry audio data are analyzed for voice content without requiring external reference signals or additional test equipment, making the system self-sufficient
3Loss of information
If all audio packets are monitored for quality assessment, then comprehensive quality data is obtained, but the processing overhead and resource consumption increase
Solution Approach 1:
The patent extracts only the essential information needed for quality assessment by identifying and focusing on packets containing actual voice content. By taking out and prioritizing voice packets over silence packets for detailed analysis, the system reduces processing overhead while maintaining information completeness where it matters most
Data Source
AI summary
Methods and systems for evaluating an audio quality using voice activity in a 4G and/or 5G network are disclosed. According to an implementation, a network component may determine that a voice call from a first user to a second user is established in a wireless network. The network component may include a user plane function in a 5G network and/or a serving gateway/packet data network gateway in a 4G network. The network component may obtain an audio stream associated with the voice call periodically until the voice call ends. The network component may further measure a packet loss rate of the audio stream, a muting rate of the audio stream, and a voice activity of the audio stream. The network component may determine whether the user perception of the voice call is impacted based at least in part on the packet loss rate, the muting rate, or the voice activity.