(A) heuristic approach to the design of a multipoint linkage teleprocessing network under capacity option

Abstract

This thesis deals with the problem of the design of the minimum-cost centralized network with multipoint linkage where options are available as to discrete link capacities. In this thesis, the concept of the ``main path`` which is the path from a node to the center is introduced. Information requests, link capacity, and link cost are considered as important factors. It is assumed that the rate and manner in which information is requested at terminals is fixed. It is shown that the problem can be formulated as a mixed integer program. Many heuristic methods as well as the optimization technique for one capacity were developed, of which Esau-Williams algorithm is known to be the best. This thesis develops a heuristic algorithm for the case of capacity options, based on Esau-Williams algorithm. This algorithm essentially applies the link exchange technique, and in each link exchange, the flow change with the associated capacity change on the main path is checked. The motivation for such efforts also stems from a need of an advanced procedure that can be applied to the case of discrete cost functions.