Survey on Qoe\Qos Correlation Models Formultimedia Services

Abstract

This paper presents a brief review of some existing correlation models which attempt to map Quality of Service (QoS)

Multivariate statistical technique over QoS variables to analyze video quality metrics on IEEE 802.11ac networks

“…Basically, they are classified in regression techniques [11] or machine learning (artificial neural networks techniques). On one hand, when we work with regression techniques, we work with all variables regardless of their inter dependences.…”

The Impact of Mobile Device Preference on the Quality of Experience

“…QoE is important for end user performance measurement to assess users' satisfaction to service providers. The parameters that affect QoE can be classified into three groups [3], Quality of video/audio content at the source, Quality of Service (QoS) which refers to the delivery of content over the network and human perception, which includes expectations and ambiance. It is very important to measure the QoE to guarantee the execution of the service agreement.…”

Challenges of future multimedia QoE monitoring for internet service providers

“…For example, instead of just investigating IP packet statistics to quantify the level of quality for an IPTV stream, RTP traffic can be taken into consideration too. These QoS-QoE models were mainly built for voice and video services, which means that they are lacking universal applicability for other services [6]. We will talk about these models in detail in the next section.…”

“…[25] While this definition is commonly accepted and used in the research domain, traditionally, Quality of Service (QoS) approaches and their associated Key Performance Indicators (KPIs) (e.g., delay or jitter) are the core of monitoring solutions, which are deployed by the industry. Research has shown that raw QoS measurements and the associated KPIs are not necessarily a reliable predictor of QoE [6,13], which is why modern QoE models focusing on perceptual characteristics of the monitored systems can help at providing a user perspective, based on measurements from the network.…”