Solving dynamic battlespace movement problems using dynamic distributed computer networks

Abstract

This thesis develops an architecture for dynamic distributed military operations research. This architecture assumes that a network of heterogeneous computing devices connects forces throughout the battlespace. Both the raw data about the battlespace and the operations research models used to analyze this data are accessible to devices on this network. The thesis designs a system using this architecture that invokes operations research network optimization algorithms to solve problems involving movement of people and equipment over dynamic road networks. A specific application is, implemented to help a medic find the nearest aid station using a shortest path algorithm. This application marshals the most current data on unit locations and road conditions (distributed across the computing network) and locates on the network an appropriate algorithm that is then used to construct a solution. The answer is returned to the user as a web page in a form appropriate for his computing device. The application is implemented with existing technologies including the Java computer language, Konig, a Java-based tool for representing networks and graphs, and Hypertext Markup Language, a format for shared information on the Internet. This system uses operations research tools to transform data into decisions in real-time or near real-time.