On the Identification and Prediction of Stalling Events to Improve QoE in Video Streaming

Abstract

Monitoring the Quality of user Experience is a challenge for video streaming services. Models for Quality of User Experience (QoE) evaluation such as the ITU-T Rec. P.1203 are very promising. Among the input data that they require are the occurrence and duration of stalling events. A stalling even5 is an interruption in the playback of multimedia content, and its negative impact on QoE is immense. Given the idiosyncrasy of this type of event, to count it and its duration is a complex task to be automated, i.e., without the participation of the user who visualizes the events or without direct access to the final device. In this work, we propose two methods to overcome these limitations in video streaming using the DASH framework. The first method is intended to detect stalling events. For simplicity, it is based on the behavior of the transport layer data and is able to classify an IP packet as belonging (or not) to a stalling event. The second method aims to predict if the next IP packet of a multimedia stream will belong to a stalling event (or not), using a recurrent neural network with a variant of the Long Short–Term Memory (LSTM). Our results show that the detection model is able to spot the occurrence of a stalling event before being experienced by the user, and the prediction model is able to forecast if the next packet will belong to a stalling event with an error rate of 10.83%, achieving an F1 score of 0.923.

A feature selection for video quality of experience modeling: A systematic literature review

“…The higher the correlation, the higher the absolute value of the weight vector. Once the subspace dimension has been set, PFA clusters highly correlated features/variables/attributes together using K‐means (Martinez‐Caro & Cano, 2021). UFS methods generally examine each feature/variable/attribute individually in order to determine the strength of the relationship of the feature/variable/attribute with the response variable using statistical tests (Shalala et al, 2018).…”

Adaptive Scalable Video Streaming (ASViS): An Advanced ABR Transmission Protocol for Optimal Video Quality

“…In response to these limitations, Netflix introduced VMAF, a video quality metric crafted to mirror human perception more accurately [25,30]. By amalgamating various video quality evaluation methods using a Support Vector Machine (SVM), VMAF produces a perceptual quality score [12,24,31]. This integrated approach ensures the holistic representation of individual measures, bolstering its correlation with subjective assessments.…”

“…To compute VMAF, spatial and temporal video quality measures are initially extracted as feature maps [12,31]. Spatial metrics are derived from the Detail Loss Metric (DLM) and Visual Information Fidelity (VIF), while the temporal metric employs Temporal Information (TI).…”

“…This stored content starts to play before the entire content is downloaded. It also helps tolerate network outages without interrupting playback by using the stored content on the local disk [12]. For clarity, the video content stored before playback on the client side is termed a Video Buffer (VB).…”

On Machine Learning Based Video QoE Estimation Across Different Networks

“…The performance of QoE/KPI estimation models is greatly affected by playback-related user interactions, which has been demonstrated in [24], stressing the need to include such interactions in the model training phase. The actual utilization of QoE/KPI estimation models in the network has been briefly addressed in [25], [26], [27], but the exact mapping of these models to network architectures and the amount of resources required for their operation is still unclear.…”