Perturbation analysis of feedback-controlled stochastic flow systems

This paper uses stochastic flow models (SFMs) for control and optimization of discrete event systems operating with a feedback control mechanism, building on earlier work that has studied such SFMs without any feedback. Using infinitesimal perturbation analysis (IPA), we derive gradient estimators for loss and workload related performance metrics with respect to threshold parameters used for buffer control. These estimators are shown to be unbiased. They are also shown to depend only on data observable from a sample path of the actual discrete event system. This renders them computable in on-line environments and easily implementable for control and performance optimization purposes.