Macroscopic Economic Planning Models for Distributed Information Systems
Raymond J. A. Buhr, C. Murray Woodside · INFOR Information Systems and Operational Research · 1977
A mathematical model for the cost of a geographically distributed information system or computer network as a function of the number of nodes is described. The cost is modelled as the sum of five separate components, due to hardware, support, user communications, computer-to-computer communications, and system unavailability. System response time is considered as a constraint, and not incorporated in the cost calculation. The model follows the general philosophy of location theory models (assuming, for instance, uniformly distributed users). However, it includes a large number of practical factors in information system design, such as vertical and horizontal distribution of resources along the dimensions of a hierarchy, economies of scale in hardware and support, operating system overhead, communication tariff structure, communications network redundancy, and the shape of the user region (ranging from a disc to a line segment, the latter to represent the Canadian situation). Five practical examples are discussed to illustrate the interpretation of the model parameters.