Perturbation analysis and optimization of multiclass multiobjective Stochastic Flow Models

Stochastic flow models (SFMs) are stochastic hybrid systems that abstract the dynamics of complex discrete event systems involving the control of sharable resources. SFMs have been used to date to study systems with a single user class or some multiclass settings in which performance metrics are not class-dependent. In this paper, we develop a SFM framework for multiple classes and class-dependent performance objectives in which we can analyze new, occasionally counterintuitive, phenomena and give rise to a new type of ¿induced¿ events that capture delays in the SFM dynamics. In the case of two classes, we derive infinitesimal perturbation analysis (IPA) estimators for their derivatives and use them as the basis for on-line optimization algorithms that apply to the underlying discrete event system (not the SFM). This allows us for the first time in the use of SFMs to contrast system-centric and user-centric objectives.