Designing and Scaling Proactive, Self-Organizing Video Servers: A Formal and a Simulation Model

Nowadays video-on-demand servers must fulfil high expectations to attract and keep clients. New concepts and architectures are being developed to improve services and take full advantage of the available resources. One promising technique is offensive adaptation, which increases the availability of insufficient resources as needed. Such servers usually need to be proactive and have flexible, self-organizing architecture. This book describes how to develop VoD servers, scale them before installation and evaluate their performance. It analyses predictor functions and similarity measures to forecast client behaviour and to achieve proactivity. A formal model is presented which facilitates the planning and evaluation of VoD servers and supports their self-organizing behaviour. Experiences gained from previous concepts are combined in a simulation environment to assist in the design of such servers and their algorithms. Finally, the results of one large and two small-scale performance studies are presented, comprising a method for performance analysis as well. Video server developers, VoD system designers and other researchers in this field should benefit from reading this book.

Peter Karpati, MSc, Dr.techn.: Computer science and mathematics at Eötvös Loránd University, Hungary; Doctoral studies in the Information Technology Institute at the University of Klagenfurt, Austria; Postdoc fellowship in the Department of Computer and Information Science at the Norwegian University of Science and Technology.