David Melendi

User behavior analysis of a video-on-demand service with a wide variety of subjects and lengths

This paper presents the analysis performed on the www.lne.es video-on-demand service (LNE TV), which is part of the digital version of one of the most important newspapers in Spain. Its principal special characteristic is the wide range of subjects (news, music, culture, tourism, nature, sports, etc) and lengths of the offered contents (from 2 minutes to 2 hours), which make it an interesting case study. Elements about user behavior have been analyzed such as session analysis; delivered time, pause distribution, jumps length, etc. The study points out interesting results about length dependence, interactions appearing, popularity of the videos, etc., which are compared with the results obtained in previous works, generally developed on educational environments (services or users). A behavior user model has been performed using the results of the analyses. This model is oriented to services with different types of information and lengths for the videos. This study has been performed thanks to an access log database with more than 150,000 requests of almost 900 videos stored over a period of 4 years. The conclusions of the study are essential to improve the service configuration and content selection. Moreover, they can be used to develop service models for video-on-demand services, which can help administrators to predict future situations and avoid performance problems.

Limitations of network emulation with single-machine and distributed ns-3

Research on large-scale internet services requires an extensive evaluation prior to deployment. A good analysis must include tests over large networks, using real devices and a considerable number of users. However, how to test in these scenarios with many users is an open question. Network emulation can be a good alternative before real deployments, which are complex and expensive. In this paper, we examine the new ns-3 network simulator/emulator in order to determine its capacity in the evaluation of large scale services. For that purpose, a real client/server video service is deployed over an emulated network. The service is progressively scaled up by increasing the number of clients on a single machine. In addition, we have extended ns-3 to support a distributed architecture for network nodes, thus, we repeat the experiments with a distributed set-up. Advantages, disadvantages, possibilities and limitations of both approaches are thoroughly discussed.

Statistical characterization of a real video on demand service: User behaviour and streaming-media workload analysis

Abstract The video on demand service characterized in this article has surpassed 4 years of service, and the access log database stores information of more than 160,000 reproductions and 900 videos. The wide variety of subjects, the range of content lengths and a frequent update of contents (several new videos every day), clearly differentiate this study from other previous research limited to specific users, subjects or environments. We present a statistical study of user behaviour and streaming traffic, analyzing session characteristics, mistaken reproductions, amount of media delivered, number and length of pauses and jumps in the reproduction, popularity and daily access profile. The results of the analysis will allow us to develop simulation models and workload generators to evaluate different scenarios and situations of the service.

Probabilistic analysis and interdependence discovery in the user interactions of a video news on demand service

Accurate workload prediction in multimedia systems requires a description of both its probabilistic properties and the interdependence between the variables of the system. This paper analyzes the interactions of the users of a news and entertainment video-on-demand service (LNE TV, http://tv.lne.es) during six months of activity. The analysis was performed using information extracted from the server logs, analyzing more than 300,000 requests for almost 1500 videos. The type of content on the site and the structure of LNE TV, which is similar to most Internet news services supported by a traditional newspaper, make this work an interesting case study, and the results can be easily extrapolated to similar audio/video streaming services. This paper shows that client interactivity in multimedia systems has the dependence structure of a stochastic process. In previous work, this type of study has been carried out in a single variable, whereas the user interactions in actual multimedia systems are multivariate. In this context, we have used copulas to create distributions to model correlated data. In our study, we model user interactions depending on their marginal univariate distributions and their correlation coefficient with the length of the videos.

Popularity analysis of a video-on-demand service with a great variety of content types: influence of the subject and video characteristics

This paper presents the popularity analysis performed on the www.lne.es video-on-demand service, which is part of the digital version of one of the most important newspapers in Spain. Its most special characteristics are the wide range of subjects (events, music, culture, tourism, nature, sports, etc) and the type of contents offered (long life, seasonal, etc), which make it an interesting case study. The analysis pays attention to the influence of the subject and the video characteristics on the popularity. Moreover, the results are checked for short and long periods of time. This study has been performed thanks to a log database with more than 170,000 requests of almost 1000 videos stored over a period of 4 years.

Adaptable system based on Scalable Video Coding for high-quality video service

Content adaptation to a heterogeneous environment like the Internet is a key process for improving the perceived quality of the user. This paper presents an adaptive streaming system using Scalable Video Coding (SVC) technology. Using feedback information from clients about the transmission status, the server is able to select the most suitable combination of SVC layers for the available bandwidth. The estimation of the available bandwidth is carried out with non-intrusive methods, based on classic metrics such as packet loss, jitter and novel metrics like the linearity of reception times of RTP packets. The system is implemented in real equipment and the results show the correct operation and the accuracy of the system when adapting to different variations of the available bandwidth. We also study the scalability of the system when several clients access the service simultaneously, demonstrating that our system is as scalable as a non-adaptive system with SVC.

Evaluation of Virtual Keyboards for Interactive Digital Television Applications

The development of the technologies behind Interactive Digital Television (IDTV) services has produced a new type of audience. Traditional viewers now become users as they may play an active role in front of the TV, for example, by choosing a video to be played on demand or by introducing text on an IDTV application. In these services, interactions need to be performed with a remote control, currently the main interaction device, or other devices such as keyboards or mice, which are not very popular in this environment. Nevertheless, although remote controls are essential tools for IDTV services, they are very limited when it comes to writing text. Thus, this study evaluated different alternatives to introduce text on an IDTV application with a remote control. A heterogeneous group of people was selected to write predefined sentences in Spanish in a test environment using three virtual keyboard layouts and the multitap mechanism. Their performance and subjective impressions reveal weaknesses and strengths of the evaluated methods. The article draws important conclusions about the usage of remote controls in IDTV applications, not only for the design of new applications but also for the research of new techniques to introduce text.

Analysis Tool for a Video-on-Demand Service Based in Streaming Technology

This paper describes the tool developed in order to analyse a video-on-demand streaming service. The aim of this work is to provide a powerful system to help both the service providers and the communication operators to configure this sort of services. Distributing the contents, developing redistribution routes, creating new contents in the most popular subjects and increasing or decreasing the length of information depending on the subscribers’ behaviour can improve the quality of the service. However, these decisions must be taken based in service performance. This tool tries to fill the gap in this field and provide the necessary analysis to configure these services. The quality of the tool has been evaluated by the www.lne.es, which is one of the most successful digital news sites in Spain. Its multimedia section offers a large number of videos on demand with several subjects, lengths and qualities. During the last months an analysis process has been performed to improve the service by using this analysis tool. This work is included in a project about analysis, modelling and configuring of interactive multimedia services.

An Empirical Investigation Into Text Input Methods for Interactive Digital Television Applications

Nowadays there is a huge market emerging in the interactive digital TV realm. In this context, we need new and effective methods of user interaction, as the main interaction device is still the classical remote control. Remote controls are especially problematic when it comes to writing text, something needed in most applications. Thus, we have carried out an empirical investigation to find effective methods of text entry with remote controls. We analyze several methods by performing experiments based on a methodology in which a heterogeneous set of real users carries out several sequential tasks in an incremental process. We analyze entry speeds, error rates, learning profiles, and subjective impressions, taking into account the particular characteristics of the users. Our results show, for instance, that Multitap is a good method for simple texts. It is between 12% and 34% faster than the fastest virtual keyboard, depending on the age of the user. Nevertheless, when complex texts need to be written, virtual keyboards present the same or even better writing speeds (QWERTY is 13% faster) and with significant lower error rates (Multitap is 347% worse than QWERTY). We consider that our results are very interesting for researchers, designers of TV applications, and hardware vendors.

Subjective evaluation of critical success factors for a QoE aware adaptive system

For many years video content delivery has established itself as the killer application. Improving QoE on adaptive streaming is focusing many efforts in the quest for optimized methods and metrics to allow a QoE driven adaptation. Questions such as whether adaptive systems based on Scalable Video Coding improve subjective quality and in which situations or to what degree are still open issues. Tolerance and indifference thresholds for each type of content, conditions or viewer category, with regard to adaptive systems are critical success factors that are yet unresolved. We compare the performance of a complete adaptive system with the traditional, i.e. non-adaptive, approach in subjective terms. Results of surveying 75 participants show that the adaptation improves QoE under most of the evaluated conditions. Tolerance thresholds for triggering adaptation events have been identified. Users accustomed to Internet video are more critical than users that only watch TV. The under 35year old subset among the available population is generally more satisfied with the adaptive system than the older subset.