The call-wise control of a communication network, such as dynamic routing, is a very complex problem, although the extent of difficulty depends on the control target. Especially, the optimization of a communication network in a multiservice environment contains the difficulty of the choice of the state variables. The analytical method is limited in handling such a problem. This paper discusses the application of the fuzzy system by extending the range of search for the optimal solution for the communication network control from the analytic single expression to the description by the combination of several rules. The usefulness of the idea is demonstrated by a simulation. Then aiming at the full utilization of the descriptive power of the fuzzy system, the modularization of the control system corresponding to network installations, as well as the hierarchical structure corresponding to the data abstract level, is proposed. Further, this paper considers the revenue maximization control of the communication network as an example, and the usefulness of the hierarchical fuzzy system is demonstrated through a simulation. By applying the proposed method, it is made possible to cope with a range of problems not previously handled by the analytical method.
Read full abstract