Team dionysos

Members
Overall Objectives
Scientific Foundations
Application Domains
Software
New Results
Contracts and Grants with Industry
Other Grants and Activities
Dissemination
Bibliography

Section: New Results

QoE (Quality of Experience)

Participants : Sebastián Basterrech, Sofiene Jelassi, Adlen Ksentini, Kandaraj Piamrat, Gerardo Rubino, Kamal Singh, César Viho.

We continue the development of the PSQA technology (Pseudo-Subjective Quality Assessment) in the area of Quality of Experience (QoE). PSQA is a method to build measuring modules capable of quantifying the quality of a video or an audio sequence, as perceived by the user, when received through an IP network. PSQA provides an accurate and efficiently computed evaluation of the quality. Accuracy means that PSQA gives values close to those than can be obtained from a panel of human observers, under a controlled subjective testing experiment, following an appropriate norm (which depends on the type of sequence or application). Efficiency means that our measuring tool can work in real time, if necessary. Observe that perceived quality is the main component of QoE.

At the heart of the PSQA approach there is the statistical learning process necessary to develop measuring modules. So far we have been using Random Neural Networks (RNNs) as our learning tool. In [58] we continue the exploration of improved versions of the main learning procedures. The general idea is to extend techniques designed for the standard Artificial Neural Network field to RNNs. In [59] we discuss how to combine the PSQA approach with the good qualities of the PESQ (Perceptual Evaluation of Speech Quality) method for assessing the quality of voice sequences. PESQ designs a set of techniques capable to evaluate the perceived quality of speech but needing access to the original signal. We follow here the “black-box” PSQA with the goal of building a PESQ-like tool but in the no-reference class, that is, no needing access to the sequence before distortion by the network. In [50] we discuss the use of PSQA-based tools in order to design new pricing techniques in the networking domain. The idea is to use the fact that PSQA allows to analyze network components and systems addressing the ultimate target from the users' point of view, the quality as they perceive it, and that way, to build pricing schemes having quality as the main target. [50] provides some preliminary ideas about this important issue for the future of networks. In [51] , we explore the case of DVB-H networks with variable bit rate H-264 coding, extending our previous work in no-reference video quality assessment.

On the other hand, deployment of heterogeneous wireless networks (i.e. 4G) is spreading throughout the world, as users want to be connected anytime, anywhere, and anyhow. Meanwhile, these users are interested more and more in multimedia applications such as video streaming and Voice over IP (VoIP), which require strict Quality of Service (QoS) support. Provisioning such constraints in this kind of system is very challenging. In fact, consider the availability of various access technologies: WiFi, WiMAX, or Cellular networks; it is difficult for a network operator to find reliable criteria to select the best network that ensures users satisfaction while maximizing network utilization. Thus, designing efficient Radio Resource Management (RRM) is mandatory for tackling such constraints. In order to provide a better understanding of RRMs design, we presented in [25] a detailed investigation of key challenges that constitutes an efficient RRM framework as well as a classification of existing solutions on RRM, in term of decision-making. Moreover, we investigated in [47] ,[78] how QoE can help for designing efficient RRM for wireless networks. In [47] we presented a QoE-based scheduler for 3G networks. Unlike the existing solution that are based on data rate or other QoS parameters, our solution propose to use PSQA (QoE metric) in order to schedule users transmission on the uplink direction. In [78] we proposed a novel network selection mechanism for heterogonous wireless networks that takes QoE into consideration for decision-making. The main idea is to use quality of experience of ongoing users in candidate networks as an indicator to select the best network for connection.


previous
next

Logo Inria