Do You Know the Average Mean Opinion Score in Your Network?

Originally published on May 07, 2018 by Sascha Neumeier
Last updated on January 23, 2024 • 9 minute read

Maybe you haven't considered your network's Mean Opinion Score yet. If you already have a Voice over IP PBX, you should keep an eye on the MOS score of your network.

What Is the MOS Score (Mean Opinion Score)?

The Mean Opinion Score (MOS) has been introduced to provide quantitative assessment of the quality of VoIP communications. Wikipedia defines the MOS score as follows:

Mean opinion score (MOS) is a measure used in the domain of Quality of Experience and telecommunications engineering, representing overall quality of a stimulus or system. It is the arithmetic mean over all individual “values on a predefined scale that a subject assigns to his opinion of the performance of a system quality”.[1] Such ratings are usually gathered in a subjective quality evaluation test, but they can also be algorithmically estimated.

MOS is a commonly used measure for video, audio, and audiovisual quality evaluation, but not restricted to those modalities. ITU-T has defined several ways of referring to a MOS in Recommendation P.800.1, depending on whether the score was obtained from audiovisual, conversational, listening, talking, or video quality tests. (Wikipedia)

How Is the MOS Score Measured?

Basically, the MOS score is somewhere between the values of 1 and 5. A MOS score of 5.0 certifies excellent voice quality. A a comparison: Public Switched Telephone Network (PSTN) has a MOS score between 4 and 5, where 5 is the absolute optimum case.

The following table provides an initial overview of the different quality levels of the Mean Opinion Score.

MOS	Quality	Impairment
5	Excellent	No effort is required to understand the language
4	Good	Through attentive listening, speech can be heard without effort
3	Fair	The language can be heard with a slight effort
2	Poor	It takes a lot of concentration and effort to understand the transmitted language
1	Bad	Despite great efforts, it is impossible to communicate

The gradation of the score depends on the measurement. While (very subjective) measurements carried out by humans can usually only be defined to one decimal place, automated measurements provide more precise results, and can be interpreted to up to 2 or 3 decimal places.

The Different MOS Metrics at a Glance

In the field of Voice over IP, the following four metrics are mainly used for validation.

Listening MOS is a forecast of the wideband Listening Quality (MOS-LQ) of the played audio stream. This value takes into consideration the audio fidelity and distortion, and predicts how a large group of users would rate the quality of the audio they hear.

Sending MOS is a forecast of the wideband MOS-LQ of the audio stream that is being sent from the user. This value takes into consideration the speech and noise levels of the user, along with any distortion, and predicts how a large group of users would rate the audio quality they hear.

Network MOS is a forecast of the wideband MOS-LQ of audio that is played. This value takes into consideration only network factors such as the codec used, packet loss, and packet errors.

Conversational MOS is a forecast of the narrowband Conversational Quality (MOS-CQ) of the audio stream that is played. This value takes into consideration the listening quality of the audio played and sent across the network, the speech and noise levels for both audio streams, and echoes. It represents how a large group of people would rate the quality of the connection for holding a conversation.

What Affects the MOS Score?

It is important to understand that the MOS score is measured on a relative scale, built on a variety of factors that can affect voice quality. Because of this, there are many factors that can affect the MOS score on VoIP systems that would not be a factor on a regular phone line. These include, among others:

Hardware
The hardware you use within your network as well as your VoIP hardware
Bandwidth
Your internal network bandwidth (limited by your network infrastructure) and your external bandwidth (limited by your leased line)
Jitter
Jitter is measuring time difference in packet inter-arrival time
Packet Loss
Packet loss is when one or more packets of data transmitted across a network fail to reach their destination
Codec Version
Various codecs are available for voice transmission. Each codec generates a certain data stream (required bandwidth), resulting in a certain MOS score.

Comparison of the Most Important Audio Codecs

Codec	Name/Description	Bandwidth	MOS
G.711	Pulse Code Modulation (PCM)	56 oder 64 kbit/s	4,10
G.722	Adaptive Multi-Rate (AMR)	48 bis 64 kBit/s	4,50
G.723.1	Multiple Maximum Likelihood Quantization (MP-MLQ)	5,6 - 6,3 kbit/s	3,90
G.723	Algebraic Code Excited Linear Prediction (A-CELP)	5,3 kbit/s	3,65
G.726	Adaptive Differential Pulse Code Modulation (AD-PCM)	16 - 40 kbit/s	3,85
G.728	Low Delay Code Excited Linear Prediction (LD-CELP)	16 kbit/s	3,61
G.729	Conjugate Structure Algebraic Code Excited Linear Prediction (CSA-CELP)	8 kbit/s	3,92
G.729A	Conjugate Structure Algebraic Code Excited Linear Prediction (CSA-CELP)	8 kbit/s	3,70

Monitoring the MOS Score in Your Network Environment

With the information gained so far, you learned that the highest possible MOS score is important for high-quality VoIP voice transmission, and that the MOS score can vary depending on the network section within the company infrastructure. It is therefore important to keep an eye on the MOS score in every corner of your network.

To prevent two admins from making test calls in the network all day long and recording the determined voice quality, there is special measuring software that carries out this process automatically. The results provided are either stored in a software database or retrieved and processed by a central network monitoring system. This allows a Quality of Service to be proven and maintained on a permanent basis.