Abstract |
Over the last years, the increasing number of mobile devices, their capabilities and the access in
wireless network have created an enormous rise on wireless traffic demand and use. Wireless networks often experience ''periods of severe impairments'', causing severe degradation to the performance of the service running on wireless devices and to the respective user experience.
However, the impact of the network performance on the quality of experience (QoE) for various
services is not understood in depth. Thus, assessing the impact of different network conditions
and system parameters on the user experience is important for improving the telecommunication
services. In general, depending on the type of service and the context, the QoE can be affected by
various techno-socio-economic-cultural-psychological parameters, e.g., by the user preferences
with respect to QoE and price, willingness - to - pay, and intrinsic indicators towards a service
provider (e.g., brand name, perceived value, reliability), its content (e.g., richness, diversity,
searching mechanisms), and even integration with other popular services (e.g., social networking
applications). In the related work, the majority of efforts aim to characterize and predict the user
experience, analyzing various types of measurements often in an aggregate manner.
Our group developed the uQoE, a modular framework that includes monitoring and data
collection tools (uQoE tracker) and algorithms for user-centric analysis and prediction of the QoE
(MLQoE prediction algorithm) in the context of video streaming service. The uQoE tracker collects
network and system measurements as well as feedback from the user. The MLQoE employs
several machine learning (ML) algorithms and tunes their hyper-parameters, given as input the
uQoE tracker collected measurements. It dynamically selects the ML algorithm that exhibits the
best performance and its parameters automatically based on the input (e.g., network and system
metrics). In this thesis, we applied the uQoE for analyzing and predicting the QoE of the video streaming service in the context of two field studies, one performed in the production
environment of a large telecom operator and the other at our Institute. The analysis indicated the
parameters with the dominant impact on the perceived QoE and revealed that the QoE may vary
across users. This motivates the use of customized adaptation mechanisms in video streaming to
address the degradation in network performance.
The MLQoE results in fairly accurate predictions.
|