Strong P-fairness: a scheduling strategy for real-time applications

Many real-time multimedia applications consume data at a steady rate over an extended period of time. This data often needs to be retrieved from storage servers, or captured at remote locations and transported over communication networks to the application. We describe a new approach to the design of such systems. We formalize a model of data usage, define a measure of temporal fairness, and present results which illustrate both the scope and the limitations of our approach.<>