Accurate Throughput Prediction (TP) represents a real challenge for reliable adaptive streaming in challenging mediums, such as cellular networks. State-of-the-art solutions adopt Deep Learning (DL) models to improve TP accuracy for various multimedia systems. This article illustrates that designing black-box TP engines that depend solely on the model’s capacity and power of learning does not achieve consistent accuracy across all throughput ranges. Additionally, we propose MATURE, a novel multistage DL-based TP model designed to capture network operating context to improve prediction accuracy. MATURE’s prediction involves characterizing the operating context before estimating the network throughput. We show that MATURE delivers consistent, accurate prediction for all throughput ranges in both 4G and 5G networks. We also show that light-weight MATURE models that use quantized parameters maintain their accuracy while featuring up to 100× faster inference, thus making them suitable for mobile implementation. Our real video streaming experiments further show that MATURE improves the average user Quality of Experience by up to 20% when compared to other TP methods.

The rise in the consumption of multimedia content has resulted in the demand to provide an exceptional user experience. However, modelling user-perceived Quality of Experience (QoE) presents a complex challenge. This fuels ongoing efforts to understand and measure QoE. To assess it, researchers rely on both subjective quality assessment (where users view and rate videos) and objective quality assessment (where quality metrics are designed to calculate perceived video quality). Although subjective evaluation is essential for mapping user experience to objective metrics, many studies omit the analysis of the impact of content preferences on user experience. This oversight limits our understanding of how video content influences QoE perception.To address this limitation, we created a multi-content video dataset with impairments based on realistic network conditions. Furthermore, we conducted a subjective study in a controlled environment evaluating the impact of user content category preferences and user video preference on QoE. One of our initial findings is that the actual video denoted as uninteresting by users had the most negative impact on the overall user QoE, but had no impact on user perception of other video degradations.

The continuous rise of multimedia entertainment has led to an increased demand for delivering outstanding user experience of multimedia content. However, modeling user-perceived Quality of Experience (QoE) is a challenging task, resulting in efforts for better understanding and measurement of user-perceived QoE. To evaluate user QoE, subjective quality assessment, where people watch and grade videos, and objective quality assessment in which videos are graded using one or many objective metrics are conducted. While there is a plethora of video databases available for subjective and objective video quality assessment, these videos are artificially infused with various temporal and spatial impairments. Videos being assessed are artificially distorted with startup delay, bitrate changes, and stalls due to rebuffering events. To conduct a more credible quality assessment, a reproduction of original user experiences while watching different types of streams on different types and quality of networks is needed. To aid current efforts in bridging the gap between the mapping of objective video QoE metrics to user experience, we developed DashReStreamer, an open source framework for re-creating adaptively streamed video in real networks. The framework takes inputs in the form of video logs captured by the client in a non-regulated setting, along with an .mpd file or a YouTube URL. The ultimate result is a video sequence that encompasses all the data extracted from the video log. DashReStreamer also calculates popular video quality metrics like PSNR, SSIM, MS-SSIM, and VMAF. Finally, DashReStreamer allows creating impaired video sequences from the popular streaming platform YouTube. As a demonstration of framework usage, we created a database of 332 realistic video clips, based on video logs collected from real mobile and wireless networks. Every video clip is supplemented with bandwidth trace and video logs used in its creation and also with objective metrics calculation reports. In addition to dataset, we performed subjective evaluation of video content, assessing its effect on overall user QoE. We believe that this dataset and framework will allow the research community to better understand the impacts of video QoE dynamics.

Accurate Throughput Prediction (TP) represents a cornerstone for reliable adaptive streaming in challenging mediums, such as cellular networks. Challenged by the highly dynamic wireless medium, recent state-of-the-art solutions adopt Deep Learning (DL) models to improve TP accuracy. However, these models perform poorly in critical, rare network conditions, leading to degraded user Quality of Experience (QoE). Such performance results from depending solely on the model's capacity and power of learning, without integrating system knowledge into the design. In this paper, we propose MATURE, a novel multi-stage DL-based TP model designed to capture network operating context to improve prediction accuracy and user experience. MATURE's operation involves characterising the operating context before estimating the network throughput. Our performance evaluation shows that MATURE improves the average user QoE by 4% - 90% in critical network conditions when compared to state-of-the-art.

Different industries are observing the positive impact of 360 video on the user experience. However, the performance of VR systems continues to fall short of customer expectations. Therefore, more research into various design elements for VR streaming systems is required. This study introduces a SW tool that offers straight-forward encoding platforms to simplify the encoding of DASH VR videos. In addition, we developed a dataset composed of 9 VR videos encoded with seven tiling configurations, four segment durations, and up to four different bitrates. A corresponding tile size dataset is also provided, which can be utilised to power network simulations or trace-driven emulations. We analysed the traffic load of various films and encoding setups using the dataset that was presented. Our research indicates that, while smaller tile sizes reduce traffic load, video decoding may require more computational power.

We consider the problem of maximum-likelihood estimation in linear models represented by factor graphs and solved via the Gaussian belief propagation algorithm. Motivated by massive Internet of Things (IoT) networks and edge computing, we set the above problem in a clustered scenario, where the factor graph is divided into clusters and assigned for processing in a distributed fashion across a number of edge computing nodes. For these scenarios, we show that an alternating Gaussian belief propagation (AGBP) algorithm that alternates between inter- and intracluster iterations, demonstrates superior performance in terms of convergence properties compared to the existing solutions in the literature. We present a comprehensive framework and introduce appropriate metrics to analyze the AGBP algorithm across a wide range of linear models characterized by symmetric and nonsymmetric, square, and rectangular matrices. We extend the analysis to the case of dynamic linear models by introducing the dynamic arrival of new data over time. Using a combination of analytical and extensive numerical results, we show the efficiency and scalability of the AGBP algorithm, making it a suitable solution for large-scale inference in massive IoT networks.

Multimedia streaming over the Internet (live and on demand) is the cornerstone of modern Internet carrying more than 60% of all traffic. With such high demand, delivering outstanding user experience is a crucial and challenging task. To evaluate user Quality of Experience (QoE) many researchers deploy subjective quality assessments where participants watch and rate videos artificially infused with various temporal and spatial impairments. To aid current efforts in bridging the gap between the mapping of objective video QoE metrics to user experience, we developed DashReStreamer, an open-source framework for re-creating adaptively streamed video in real networks. DashReStreamer utilises a log created by a HTTP adaptive streaming (HAS) algorithm run in an uncontrolled environment (i.e., wired or wireless networks), encoding visual changes and stall events in one video file. These videos are applicable for subjective QoE evaluation mimicking realistic network conditions.

Evolving Internet applications, such as immersive multimedia and Industry 4, exhibit stringent delay, loss, and rate requirements. Realizing these requirements would be difficult without advanced dynamic traffic management solutions that leverage state-of-the-art technologies, such as Software-Defined Networking (SDN). Mininet represents a common choice for evaluating SDN solutions in a single machine. However, Mininet lacks the ability to emulate links that have multiple queues to enable differentiated service for different traffic streams. Additionally, performing a scalable emulation in Mininet would not be possible without light-weight application emulators. In this paper, we introduce two tools, namely: QLink and SPEED. QLink extends Mininet API to enable emulating links with multiple queues to differentiate between different traffic streams. SPEED represents a light-weight web traffic emulation tool that enables scalable HTTP traffic simulation in Mininet. Our performance evaluation shows that SPEED enables scalable emulation of HTTP traffic in Mininet. Additionally, we demo the benefits of using QLink to isolate three different applications (voice, web, and video) in a network bottleneck for numerous users.

Fifth Generation (5G) networks provide high throughput and low delay, contributing to enhanced Quality of Experience (QoE) expectations. The exponential growth of multimedia traffic pose dichotomic challenges to simultaneously satisfy network operators, service providers, and end-user expectations. Building QoE-aware networks that provide run-time mechanisms to satisfy end-users' expectations while the end-to-end network Quality of Service (QoS) varies is challenging, and motivates many ongoing research efforts. The contribution of this work is twofold. Firstly, we present a reproducible data-driven framework with a series of pre-installed Dynamic Adaptive Streaming over HTTP (DASH) tools to analyse state-of-art Adaptive Bitrate Streaming (ABS) algorithms by varying key QoS parameters in static and mobility scenarios. Secondly, we introduce an interactive Jupyter notebook and Binder service providing a live analytical environment, which processes the output dataset of the framework and compares the relationship of five QoE models, three QoS parameters (RTT, throughput, packets), and seven different video KPIs.

In this paper, we present a 5G trace dataset collected from a major Irish mobile operator. The dataset is generated from two mobility patterns (static and car), and across two application patterns (video streaming and file download). The dataset is composed of client-side cellular key performance indicators (KPIs) comprised of channel-related metrics, context-related metrics, cell-related metrics and throughput information. These metrics are generated from a well-known non-rooted Android network monitoring application, G-NetTrack Pro. To the best of our knowledge, this is the first publicly available dataset that contains throughput, channel and context information for 5G networks. To supplement our real-time 5G production network dataset, we also provide a 5G large scale multi-cell ns-3 simulation framework. The availability of the 5G/mmwave module for the ns-3 mmwave network simulator provides an opportunity to improve our understanding of the dynamic reasoning for adaptive clients in 5G multi-cell wireless scenarios. The purpose of our framework is to provide additional information (such as competing metrics for users connected to the same cell), thus providing otherwise unavailable information about the base station (eNodeB or eNB) environment and scheduling principle, to end user. Our framework permits other researchers to investigate this interaction through the generation of their own synthetic datasets.

In this short paper, we present goDASH, an infrastructure for headless streaming of HTTP adaptive streaming (HAS) video content, implemented in the language golang, an open-source programming language supported by Google. goDASH's main functionality is the ability to stream HAS content without decoding actual video (headless player). This results in low memory requirements and the ability to run multiple players in a large-scale-based evaluation setup. goDASH comes complete with numerous state-of-the-art HAS algorithms, and is fully written in the Google golang language, which simplifies the implementation of new adaptation algorithms and functions. goDASH supports two transportation protocols Transmission Control Protocol (TCP) and Quick UDP Internet Connections (QUIC). The QUIC protocol is a relatively new protocol with the promise of performance improvement over the widely used TCP. We believe that goDASH is the first emulation-based HAS player that supports QUIC. The main limitation in using QUIC protocol is the need for a security certificate setup on both ends (client and server) as QUIC demands an encrypted connection. This limitation is eased by providing our own testbed framework, known as goDASHbed. This framework uses a virtual environment to serve video content locally (which allows setting security certificates) through the Mininet virtual emulation tool. As part of Mininet, goDASH can be used in conjunction with other traffic generators.

The growth of online video-on-demand consumption continues unabated. Existing heuristic-based adaptive bit-rate (ABR) selection algorithms are typically designed to optimise video quality within a very narrow context. This may lead to video streaming providers implementing different ABR algorithms/players, based on a network connection, device capabilities, video content, etc., in order to serve the multitude of their users' streaming requirements. In this paper, we present SMASH: a Supervised Machine learning approach to Adaptive Streaming over HTTP, which takes a tentative step towards the goal of a one-size-fits-all approach to ABR. We utilise the streaming output from the adaptation logic of nine ABR algorithms across a variety of streaming scenarios (generating nearly one million records) and design a machine learning model, using systematically selected features, to predict the optimal choice of the bitrate of the next video segment to download. Our evaluation results show that SMASH guarantees a high QoE with consistent performance across a variety of streaming contexts.

In this short demo paper, we introduce godash 2.0 godash is a headless HTTP adaptive streaming (HAS) video streaming platform written in the Google programming language GO. godash has been extensively rewritten for this release so as to provide ease of use, and a host of new features. godash includes options for eight different state of the art adaptive algorithms, five HAS profiles, four video codecs, the ability to stream audio and video segments, two transport protocols (TCP and QUIC), realtime output from five Quality of Experience (QoE) models, as well as a collaborative framework for the evaluation of cooperative HAS streaming. In this demo, we will introduce each of the options available in the godash configuration file, and illustrate examples of how to use the collaborative players using both godash on a native machine as well as godash within godashbed. godashbed is an integrated large-scale testbed framework for the evaluation of HAS streaming, which uses a virtual environment to serve video content locally (which allows setting security certificates) through the Mininet virtual emulation tool. In this manner, godash provides a framework for rapid deployment and testing of new HAS algorithms, QoE models and transport protocols.

Recent years have witnessed an explosion of multimedia traffic carried over the Internet. Video-on-demand and live streaming services are the most dominant services. To ensure growth, many streaming providers have invested considerable time and effort to keep pace with ever-increasing users' demand for better quality and stall abolition. HTTP adaptive streaming (HAS) algorithms are at the core of every major streaming provider service. Recent years have seen sustained development in HAS algorithms. Currently, to evaluate their proposed solutions, researchers need to create a framework and numerous state-of-the-art algorithms. Often, these frameworks lack flexibility and scalability, covering only a limited set of scenarios. To fill this gap, in this paper we propose DASHbed, a highly customizable real-time framework for testing HAS algorithms in a wireless environment. Due to its low memory requirement, DASHbed offers a means of running large-scale experiments with a hundred competing players. Finally, we supplement the proposed framework with a dataset consisting of results for five HAS algorithms tested in various evaluated scenarios. The dataset showcases the abilities of DASHbed and presents the adaptation metrics per segment in the generated content (such as switches, buffer-level, P. 1203.1 values, delivery rate, stall duration, etc.), which can be used as a baseline when researchers compare the output of their proposed algorithm against the state-of-the-art algorithms.

Today's HTTP adaptive streaming applications are designed to provide high levels of Quality of Experience (QoE) across a wide range of network conditions. The adaptation logic in these applications typically needs an estimate of the future network bandwidth for quality decisions. This estimation, however, is challenging in cellular networks because of the inherent variability of bandwidth and latency due to factors like signal fading, variable load, and user mobility. In this paper, we exploit machine learning (ML) techniques on a range of radio channel metrics and throughput measurements from a commercial cellular network to improve the estimation accuracy and hence, streaming quality. We propose a novel summarization approach for input raw data samples. This approach reduces the 90th percentile of absolute prediction error from 54% to 13%. We evaluate our prediction engine in a trace-driven controlled lab environment using a popular Android video player (ExoPlayer) running on a stock mobile device and also validate it in the commercial cellular network. Our results show that the three tested adaptation algorithms register improvement across all QoE metrics when using prediction, with stall reduction up to 85% and bitrate switching reduction up to 40%, while maintaining or improving video quality. Finally, prediction improves the video QoE score by up to 33%.

In this paper, we present a 4G trace dataset composed of client-side cellular key performance indicators (KPIs) collected from two major Irish mobile operators, across different mobility patterns (static, pedestrian, car, bus and train). The 4G trace dataset contains 135 traces, with an average duration of fifteen minutes per trace, with viewable throughput ranging from 0 to 173 Mbit/s at a granularity of one sample per second. Our traces are generated from a well-known non-rooted Android network monitoring application, G-NetTrack Pro. This tool enables capturing various channel related KPIs, context-related metrics, downlink and uplink throughput, and also cell-related information. To the best of our knowledge, this is the first publicly available dataset that contains throughput, channel and context information for 4G networks.

Streaming over the wireless channel is challenging due to rapid fluctuations in available throughput. Encouraged by recent advances in cellular throughput prediction based on radio link metrics, we examine the impact on Quality of Experience (QoE) when using prediction within existing algorithms based on the DASH standard. By design, DASH algorithms estimate available throughput at the application level from chunk rates and then apply some averaging function. We investigate alternatives for modifying these algorithms, by providing the algorithms direct predictions in place of estimates or feeding predictions in place of measurement samples. In addition, we explore different prediction horizons going from one to three chunk durations. Furthermore, we induce different levels of error to ideal prediction values to analyse deterioration in user QoE as a function of average error.

Dynamic adaptive streaming over HTTP (DASH) is widely adopted for video transport by major content providers. However, the inherent high variability in both encoded video and network rates represents a key challenge for designing efficient adaptation algorithms. Accommodating such variability in the adaptation logic design is essential for achieving a high user quality of Experience (QoE). In this paper, we present ARBITER+ as a novel adaptation algorithm for DASH. ARBITER+ integrates different components that are designed to ensure a high video QoE while accommodating inherent system variabilities. These components include a tunable adaptive target rate estimator, hybrid throughput sampling, controlled switching, and short-term actual video rate tracking. We extensively evaluate the streaming performance using real video and cellular network traces. We show that ARBITER+ components work in harmony to balance temporal and visual QoE aspects. Additionally, we show that ARBITER+ enjoys a noticeable QoE margin in comparison to state-of-the-art adaptation approaches in various operating conditions. Furthermore, we show that ARBITER+ also achieves the best application-level fairness when a group of mobile video clients shares a cellular base station.

In this demonstration we present a platform that encompasses all of the components required to realistically evaluate the performance of Dynamic Adaptive Streaming over HTTP (DASH) over a real-time NS-3 simulated network. Our platform consists of a network-attached storage server with DASH video clips and a simulated LTE network which utilises the NS-3 LTE module provided by the LENA project. We stream to clients running an open-source player with a choice of adaptation algorithms. By providing a user interface that offers user parametrisation to modify both client and LTE settings, we can view the evaluated results of real-time interactions between the network and the clients. Of special interest is that our platform streams actual video clips to real video clients in real-time over a simulated LTE network, allowing reproducible experiments and easy modification of LTE and client parameters. The demonstration showcases how changes in LTE network settings (fading model, scheduler, client distance from eNB, etc.), as well as video-related decisions at the clients (streaming algorithm, quality selection, clip selection, etc.), can impact on the delivery and achievable quality.