By Herbert J. Bernstein

© Copyright 2000,2002 Herbert J. Bernstein

Network Modeling is an aspect of information system modeling which focuses on the relationships of geography to flows of information (and flows of control) through a system. While the design of a network is most naturally graphical, tabular breakdowns of portions of the design can be very useful. The major differences among network modelling and data modeling and process modeling are that in the first we concentrate on where, when and how much information is moved, while in the second we focus on on what information is moved and where it is moved, while third the later we focus on how it is moved and when it is moved.

• The Network Model
  Physical vs. Logical
  • What information is moved where, when and at what rates
  • Based on directed graphs -- nodes (locations) and edges (connections)
  • Based on queues -- customers and servers
• Tabular Representations
  • Incidence Matrix -- what is connected to what
  • CRUD Matrices -- Create, Read (or use), Update (or write), Delete
    • Data (Entities and Attributes) to Process CRUD Matrix
      The levels of access required by each process to each data item
    • Data to Location CRUD Matrix
      The levels of access required by each location
  • Process to Location Matrix
    For each process, the locations at which that process is peformed
• Design Paradigms
  • Saturated queues grow to infinity -- reserve 20-50%
  • Flows are limited by the narrowest bottleneck