A fuzzy approach to the flow control problem

Abstract

We consider queueing systems in which arriving customers may be accepted or rejected. The objective is to determine the optimal admission policies so as to maximize the average profit (reward minus cost) over an infinite horizon. In this paper we use fuzzy logic to solve two problems for which analytical solutions exist. Fuzzy logic produces identical policies to the analytical ones and, therefore, these problems serve as benchmarks for further work. We then tackle a new problem with two arrival streams and parallel servers. The fuzzy approach is tested with simulation, and appears to be efficient and promising in areas where analysis is limited.