Application of Deep Learning Techniques to Video QoE Prediction in Smartphones

Cardenas-Angelat, C.; Polglase, Janie Banos; Vaca-Rubio, Cristian J.; Aguayo-Torres, Mari Carmen

doi:10.1109/eucnc.2019.8801974

Cited by 2 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given the analysis of the 802.11a/g/n/ac/ax standards, this study was able to achieve higher accuracy in predicting the quality of YouTube streaming compared to the results achieved in the literature [15][16][17][18][19]. While these related works used Wi-Fi as an access network for their experiments, they did not leverage 802.11 specific performance parameters for predicting the PQoS of YouTube streaming.…”

Section: Discussionmentioning

confidence: 91%

“…While these related works used Wi-Fi as an access network for their experiments, they did not leverage 802.11 specific performance parameters for predicting the PQoS of YouTube streaming. However, this study produced multiple large datasets compared to the work in References [ 15 , 16 , 18 , 19 ] and was able to achieve higher accuracy regarding the majority class prediction as the baseline. Indeed, some machine learning approaches presented in Related Works could not achieve significant improvement compared to the majority class prediction baseline approach.…”

Section: Discussionmentioning

confidence: 99%

“…The dataset comprised 1303 instances and they trained neural networks with 80% of the dataset. They were able to achieve around 90% accuracy in predicting binary classes of YouTube streaming performance parameters [ 16 ]. In other research, Bartolec et al analyzed YouTube performance using ML on Wi-Fi networks and collected the startup delay, stalling, resolution and average bitrate using YouTube stats-for-nerds on two mobile phones and emulated the LTE network condition using the IMUNES tool.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Estimating PQoS of Video Streaming on Wi-Fi Networks Using Machine Learning

Morshedi

Noll

2021

Sensors

View full text Add to dashboard Cite

Video on demand (VoD) services such as YouTube have generated considerable volumes of Internet traffic in homes and buildings in recent years. While Internet service providers deploy fiber and recent wireless technologies such as 802.11ax to support high bandwidth requirement, the best-effort nature of 802.11 networks and variable wireless medium conditions hinder users from experiencing maximum quality during video streaming. Hence, Internet service providers (ISPs) have an interest in monitoring the perceived quality of service (PQoS) in customer premises in order to avoid customer dissatisfaction and churn. Since existing approaches for estimating PQoS or quality of experience (QoE) requires external measurement of generic network performance parameters, this paper presents a novel approach to estimate the PQoS of video streaming using only 802.11 specific network performance parameters collected from wireless access points. This study produced datasets comprising 802.11n/ac/ax specific network performance parameters labelled with PQoS in the form of mean opinion scores (MOS) to train machine learning algorithms. As a result, we achieved as many as 93–99% classification accuracy in estimating PQoS by monitoring only 802.11 parameters on off-the-shelf Wi-Fi access points. Furthermore, the 802.11 parameters used in the machine learning model were analyzed to identify the cause of quality degradation detected on the Wi-Fi networks. Finally, ISPs can utilize the results of this study to provide predictable and measurable wireless quality by implementing non-intrusive monitoring of customers’ perceived quality. In addition, this approach reduces customers’ privacy concerns while reducing the operational cost of analytics for ISPs.

show abstract

Section: Discussionmentioning

confidence: 91%

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Estimating PQoS of Video Streaming on Wi-Fi Networks Using Machine Learning

Morshedi

Noll

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…For instance, video streaming QoE is modeled using metrics of visual quality of the rendered frames (e.g., SSIM [8]), quality stability (e.g., number of bitrate switches), and smoothness (e.g., rebuffering events [22]). The quality functions range from linear combinations of these metrics [45] to deep learning models [20]. Similarly, web QoE is modeled by variants of page load time (e.g., time-to-first-byte) to capture the impact of object loading progress on user QoE [17,18,21].…”

Section: Today: Quality Metrics As Featuresmentioning

confidence: 99%

“…Indeed, the march of QoE feature engineering is in full swing, fueled by new applications (e.g., interactive live video) on new devices (e.g., panoramic headsets) with new application behaviors (e.g., new online advertising methods). For example, recent work predicts video QoE using 275 features [42] and complex machine-learning models [20]. But the improvement in performance is not commensurate with this complexity.…”

Section: Introductionmentioning

confidence: 99%

A New Abstraction for Internet QoE Optimization

Jiang,

Sen

2020

Preprint

View full text Add to dashboard Cite

A perennial quest in networking research is how to achieve higher quality of experience (QoE) for users without incurring more resources. This work revisits an important yet often overlooked piece of the puzzle: what should the QoE abstraction be? A QoE abstraction is a representation of application quality that describes how decisions affect QoE.The conventional wisdom has relied on developing handcrafted quality metrics (e.g., video rebuffering events, web page loading time) that are specialized to each application, content, and setting. We argue that in many cases, it maybe fundamentally hard to capture a user's perception of quality using a list of handcrafted metrics, and that expanding the metric list may lead to unnecessary complexity in the QoE model without a commensurate gain. Instead, we advocate for a new approach based on a new QoE abstraction called visual rendering. Rather than a list of metrics, we model the process of quality perception as a user watching a continuous "video" (visual rendering) of all the pixels on their screen. The key advantage of visual rendering is that it captures the full experience of a user with the same abstraction for all applications. This new abstraction opens new opportunities (e.g., the possibility of end-to-end deep learning models that infer QoE directly from a visual rendering) but it also gives rise to new research challenges (e.g., how to emulate the effect on visual rendering of an application decision). This paper makes the case for visual rendering as a unifying abstraction for Internet QoE and outlines a new research agenda to unleash its opportunities.

show abstract

Application of Deep Learning Techniques to Video QoE Prediction in Smartphones

Cited by 2 publications

References 10 publications

Estimating PQoS of Video Streaming on Wi-Fi Networks Using Machine Learning

Estimating PQoS of Video Streaming on Wi-Fi Networks Using Machine Learning

A New Abstraction for Internet QoE Optimization

Contact Info

Product

Resources

About