Using Genetic Algorithms to Aid in a Vulnerability Analysis of National Missile Defense Simulation Software

Abstract

The National Strategy for Homeland Defense, published by the then U.S. Office of Homeland Security in July 2002, directs all U.S. government agencies to conduct vulnerability analyses of their sensitive systems. This policy applies to Department of Defense systems, including the simulation packages used to design and exercise national missile defense. Many, if not most, of the previous vulnerability analyses of missile defense simulation platforms have utilized traditional reverse engineering techniques along with a review of all documentation and publicly available sources. These experiments have produced some useful information, though the amount of platform specific data recovered has been limited. The use of genetic algorithms (GAs) has been shown to be an effective method of performing boundary analysis and parameter optimization. In this paper, we show how GAs can be used to extract information concerning how particular parameters affect heavily parameterized missile defense simulation system performance. This information would be very valuable to researchers as part of a greater vulnerability analysis of national missile defense software.