Reliability modeling and evaluation in real-time distributed multimedia systems

This paper investigates the reliability modeling and evaluation in real-time distributed multimedia systems with series, parallel, k-out-of-n, and hybrid systems. Both combinatorial and Markov modeling techniques are used. The Multimedia Digital Distribution System (MDDS), developed at Thomson-CSF Sextant In-Flight Systems, is used as a case study in which several reliability and availability models will be built. These models will then be solved to determine the mean-time to failure (MTTF), reliability, and availability of the MDDS and its subsystems. Data redundancy design with an associated reliability model for improving the MDDS reliability will also be presented. The paper is concluded with a reliability model from the point of a passenger who uses the MDDS, and a reliability model at a system level. These models are very useful in estimating the revenue for the customer airlines. Solutions of these models show that we can determine the financial value when increasing the reliability, and also show the areas of the MDDS that need to be highly reliable in order to maximize the total revenue from each airplane.