Voice Quality

Description

Voice Quality (VQ) in 3GPP refers to the technical characterization and measurement of the auditory experience delivered to a user during a voice call. It encompasses both subjective listening quality (how a human perceives it) and objective algorithmic predictions of that quality. The primary framework is defined in the P.800 series from ITU-T, which 3GPP adopts and extends for cellular-specific contexts. Key objective metrics include the Perceptual Evaluation of Speech Quality (PESQ), defined in ITU-T P.862, and its successor, the Perceptual Objective Listening Quality Assessment (POLQA), defined in ITU-T P.863. These algorithms analyze a degraded speech signal received at the listener's end against the original reference signal to predict a Mean Opinion Score (MOS), typically on a scale from 1 (bad) to 5 (excellent).

3GPP specifications define how these measurements are integrated into the network architecture. This can involve active testing, where test calls are made with known reference speech samples, or passive monitoring, where live traffic is analyzed without injecting test signals. Network elements like the Media Resource Function (MRF) or dedicated probes can perform these measurements. For VoLTE and VoNR, VQ monitoring is tightly coupled with the IP Multimedia Subsystem (IMS) and the Packet Data Convergence Protocol (PDCP) layer to correlate quality metrics with specific radio conditions, handovers, or codec changes. Parameters such as codec type (e.g., AMR, EVS), packet loss, jitter, delay, and echo all critically impact the calculated VQ score.

The role of VQ is integral to Quality of Service (QoS) and Quality of Experience (QoE) management. It provides operators with quantifiable data to perform root cause analysis for poor voice performance, whether the issue originates in the radio access network (e.g., poor coverage, interference), the core network (e.g., congestion, transcoding), or the device itself. By defining standardized measurement techniques and reporting formats, 3GPP ensures that VQ assessments are consistent and comparable across different networks and equipment vendors, which is essential for benchmarking, service level agreement (SLA) verification, and driving continuous network optimization.

Purpose & Motivation

The standardization of Voice Quality metrics was motivated by the fundamental need for mobile network operators to objectively manage and guarantee the performance of their most critical service: voice calls. As networks evolved from circuit-switched 2G/3G to packet-switched VoLTE and VoNR, the transmission became vulnerable to new impairments like IP packet loss, variable delay (jitter), and transcoding artifacts. Subjective listening tests alone are expensive, slow, and not scalable for network-wide monitoring. Therefore, objective, automated measurement tools were necessary to provide real-time or near-real-time insights into service quality.

Historically, voice quality was often inferred from radio metrics like signal strength (RxLev) or bit error rate (BER). However, these are indirect indicators and do not capture the end-to-end perceptual experience, which is affected by the entire chain from codec selection to network jitter. The introduction of standardized VQ methodologies in 3GPP, starting in Rel-8 alongside the early work on LTE and IMS, addressed this gap. It provided a common language and technical basis for evaluating performance, enabling operators to proactively identify degradation, correlate it with network events, and implement corrective actions. This was crucial for the successful introduction and adoption of VoLTE, as it had to meet or exceed the quality expectations set by traditional circuit-switched voice.

Key Features

• Standardized objective quality metrics (e.g., PESQ, POLQA MOS predictions)
• Support for both active testing and passive monitoring methodologies
• Integration with IMS and EPC/5GC for VoLTE/VoNR quality assurance
• Correlation of voice quality scores with radio and transport layer parameters
• Defined reporting formats and interfaces for network management systems
• Assessment of wideband and super-wideband speech codecs (e.g., EVS)

Evolution Across Releases

Defining Specifications