AT&T Labs Research — Leading Invention, Driving Innovation

(c) ACM, 2011. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in IMC 2011, 2011-11-02.

{We investigate how consumers view content using Video on Demand (VoD) in the context of an IP-based video distribution environment. Users today can use advanced stream control functions like skip and replay in addition to play, fast-forward, rewind, pause etc., to interactively control their viewing. Such stream control, however, places additional demands on the distribution infrastructure (servers, network, and set top boxes) and can be challenging to manage with a large subscriber base. A model of user-interaction is useful to provide key insights on their impact on server and bandwidth requirements, client responsiveness, etc. We capture user activity in their natural setting of viewing video at home. We first develop a model for the arrival process of requests for content. We then develop two stream control models that accurately capture user interaction. We show that stream control events can be characterized by a finite state machine and a sojourn time model, parametrized for major periods of usage (weekend and weekday). Our semi-Markov (SM)model for the sojourn time in each stream control state uses a novel technique based on a polynomial fit to the logarithm of the Inverse CDF. A second Constrained model (CM) uses a stick-breaking approach familiar in machine learning to model the individual state sojourn time distributions. The SM model seeks to preserve the sojourn time distribution for each state while the CM model puts a greater emphasis on preserving the overall session duration distribution. Using traces across a period of 2 years from a large-scale operational IPTV environment we validate the proposed model and show that we are able to faithfully predict the workload presented to a video server. We also provide a synthetic trace developed from the model enabling researchers to also study other problems of interest.}

The definitive version was published in COMSNETS 2011. , 2011-01-05, http://www.comsnets.org/

{Service providers are evolving to provide more video content on-demand. Customers like to watch a variety of entertainment content of their choice and at a time conducive to their schedules. Catering to this ever-increasing user base requires careful provisioning by the providers to accommodate for both scale and interactivity. In this paper, we examine the usage pattern of several hundreds of thousands of consumers of a nationwide IPTV service, and confirm that viewers are indeed migrating to what is called �time-shifted� viewing of television programming and movies using digital video recorders or on demand viewing. We also show how users of on-demand content interactively control their viewing experience using �stream control� functions such as fast-forward, rewind, skip, replay, etc. Through careful measurements on an IPTV server, we compute the load due to streaming and handling these stream control events. We then extrapolate from these micro-benchmark measurement to predict the processing load imposed by users that would resort to using a �network-based� DVR capability if such a service were offered. We use both detailed trace-driven simulations and a simple operational-analysis based model to predict the capacity requirements of the server complex in the VHO to serve a large population of customers (e.g. a densely populated city like Mumbai). We provide insights on the number of requests serviced by the server, the average time to service these requests and the response time as perceived by the client.}

{Interactivity is promised by IP-based content distribution, particularly with IPTV. We investigate the user viewing activity for broadcast TV, pre-recorded content using Digital Video Recording (DVR) and video on demand (VoD). Advanced stream control functions (play, pause, skip, rewind, etc.) provide users with a high level of interactivity, but place demands on the distribution infrastructure (servers, network, home-network) that can be difficult to manage at large scale. To support system design as well as network capacity planning, it is necessary to have a good model of user interaction. Using traces from a well-provisioned operational environment with a large user population, we first characterize interactivity for broadcast TV, DVR and VoD. We then develop parametric models of individual users stream control operations for VoD. Our analysis shows that interactive behavior is adequately characterized by two semi-Markov models, one for weekdays and another for weekends. We propose a parametric model for the underlying sojourn time distributions and show that it results in a superior fit compared to well known distributions (generalized Pareto and Weibull). In order to validate that our models faithfully capture user behavior, we compare the workload that a VoD server experiences in response to actual traces and synthetic data generated from our proposed models. }

Awards
American Statistical Association Fellow, 2005.