Design
and Analysis of Differentiated Services: Approximate
Solutions and Scaling Relations
We describe a framework for analyzing a large class of service systems that deliver differentiated services, examples of which include internet based applications, customer contact centers and rental systems. The unifying theme in our approach is the concept of stochastically modulated capacity, and a corresponding set of diffusion approximations that is driven by large capacity asymptotics. We illustrate how these methods may be used in the design, performance analysis and economic optimization of such stochastic service systems.