A Distributed Processing Network (DPN) can assume many different physical topologies. However, when a network is viewed as a group of logical connections, only a few distinct topologies are possible. They are centralized, star, and fully connected. A fourth 'hybrid' topology which combines features of two or more of these is also possible. Each of these topolo gies, or structures, may have a unique cost to the firm as well as unique benefits.
This paper presents a network cost function for each of the possible logical network structures. These functions provide the basis for an interactive model which can be used to develop cost estimates. These estimates can then be used with a sep arate analysis of benefits to aid in network design or can be used to assist in budget preparation when configuration changes are planned. A bank case study is provided to illustrate the use of the model based on these cost functions.