An intelligent sampling framework for controlled experimentation and QoE modeling

Abstract

For internet applications, measuring, modeling and predicting the quality experienced by end users as a function of network conditions is challenging. A common approach for building application specific Quality of Experience (QoE) models is to rely on controlled experimentation. For accurate QoE modeling, this approach can result in a large number of experiments to carry out because of the multiplicity of the network features, their large span (e.g., bandwidth, delay) and the time needed to setup the experiments themselves. However, most often, the space of network features in which experimentations are carried out shows a high degree of similarity in the training labels of QoE. This similarity, difficult to predict beforehand, amplifies the training cost with little or no improvement in QoE modeling accuracy. So, in this paper, we aim to exploit this similarity, and propose a methodology based on active learning, to sample the experimental space intelligently, so that the training cost of experimentation is reduced. We validate our approach for the case of YouTube video streaming QoE modeling from out-of-band network performance measurements, and perform a rigorous analysis of our approach to quantify the gain of active sampling over uniform sampling.

When Deep Learning meets Web Measurements to infer Network Performance

“…Unlike the previous class, the approaches of this class, since they do not necessarily know the network conditions that users may face, must widely explore the different possible network conditions. Several methods have been proposed to effectively explore the very wide space of possible conditions, such as the quasi-Monte Carlo method [18], the active learning method [19] or the Fourier Amplitude Sensitivity Test (FAST) [20] that we are exploring in this paper.…”

On Accounting for Screen Resolution in Adaptive Video Streaming: A QoE-Driven Bandwidth Sharing Framework

“…As per prior subjective studies, the QoE of video streaming is a function of application layer QoS features that are either dependent on the video content (e.g., video bit rate) or the playout metrics (e.g., the intial startup delay) [8], [18]; the playout metrics further depend on the underlying network conditions such as the network throughput or delay.…”

“…We also compare the different implementations in terms of main application-level QoS factors (e.g., stalls, resolution switches and interruptions) that could impact the subjective QoE [8], [18]. We focus on the quality switches as they occur only in adaptive video streaming sessions making their evaluation important.…”

“…Even though 5G networks promise high connectivity and huge transmission capacity aiming to take users experience to the next level [5], [6], bandwidth sharing is still an important issue for network operators and content providers, especially in view of the exponential rise of video traffic volume. It turns out that the objective aspect of Quality of Experience is tightly correlated to terminal playout characteristics (e.g, size, resolution) but also to network conditions [1], [7], [8].…”

On accounting for screen resolution in adaptive video streaming: QoE‐driven bandwidth sharing framework

“…Even though 5G networks promise high connectivity and huge transmission capacity aiming to take the internet services and the corresponding user experience to the next level [3], [4], bandwidth sharing is still an important issue for network operators and content providers, especially in view of the exponential rise of video traffic volume ( Figure 1). It turns out that the objective aspect of Quality of Experience is tightly correlated to terminal playout characteristics (e.g., size, resolution) but also to network conditions [5], [6], [7].…”