Understanding the quality of experience in modern distributed interactive multimedia applications in presence of failures

Abstract

Recent technological advances have made it possible to design bandwidth demanding distributed interactive multimedia applications such as the World Opera application. In this application artists from different opera houses across the globe, can participate in a single united performance and interact almost as if they were co-located.One of the main design challenges in this application domain is to determine the composition of system components necessary to satisfy the desired quality of service in presence of failures and budget constraints. This challenge is exacerbated by the fact that quality of service depends on a multitude of factors such as human perception of video and audio, the type of audience, performance elements, etc. These factors cannot be captured by traditional approaches for dependability evaluation such as reliability, i.e., continuous delivery of correct service. This calls for developing a more comprehensive "Quality of Experience" concept.In this paper, we propose a novel method to assess the quality of experience in presence of failures, based on a new metric called perceived reliability. This method can help the system designers and engineers compare architectural variants and to determine the dependability budget. We show the feasibility of our method by applying it to a World Opera performance. Our experimental results provide useful guidelines for system engineers towards improving the quality of experience of World Opera performances despite presence of failures.

Modeling QoE in Dependable Tele-Immersive Applications: A Case Study of World Opera

“…For example, in a distributed dance performance, where the audience is immersed in the video, the audience may not notice a small change in audio quality. Moreover, providing a performance that is failure-free in the classical sense is notoriously difficult and unlikely [20]. Fortunately, the audience might be able to tolerate a mild failure even if the failure is perceived, if a higher quality level is restored within a reasonable amount of time.…”

A DSL-Supported Workflow for the Automated Assembly of Large Stochastic Models

“…In such approaches (e.g., see [10,20,31,37]) the overall analysis model is built out of well-defined submodels addressing specific aspects of the systems, which are then composed following predefined rules based on the actual scenario to be represented. When submodels are reused multiple times, this approach leads to a modeling paradigm that resembles object-oriented programming (OOP): libraries of "template" submodels are created for a given system, having fixed "interfaces" and "parameters".…”

“…These Figure 1. System architecture of a World Opera stage [37,38] generated streams collectively represent the real-world data.…”

“…A set of metrics for the evaluation of the WO system have been defined in [37], and evaluated on a simplified case study. A more complex and faithful model of the system is being developed using a modular modeling approach as discussed before, in which the model is subdivided in well-specified "template" submodels, which are then instantiated multiple times and connected together following precise rules [27].…”

“…The actual implementation of these models depends on the behavior to be modeled; different implementations may for example be needed for representing different fault-tolerance mechanisms. For simplicity, we have assumed a hot-spare behavior for all components as in [37], thus leading to a single "Component" template. In case of components with different behavior, different templates would be needed.…”