IED dataset generation: Analysis across theaters

Abstract

Improvised Explosive Devices are the most dangerous threat to United States troops in Iraq and Afghanistan. Regardless of where troops deploy next, there is a strong likelihood of IED use in future conflicts. While great progress in analysis and exploitation of IED data has been made, the data is generally limited to historical attacks in current theaters of operation. The purpose of this analysis is to develop a tool that generates IED placement data based on statistically valid assumptions and distributions in any theater. These distributions can be used to train analysts prior to deployment to new environments. The research will generate data representing hypothetical IED attacks in a table structured like the CIDNE database. IED data from the CIDNE database will be used to develop distributions for each attack detailing IED location and type (type includes: command detonated, victim detonated, timer operated and radio controlled). Hypotheses will be tested using CIDNE data to determine some trends, including: seasonality, distance from road and location relative to terrain features. Results of these tests will shape the distributions used in the generation of synthetic datasets. The datasets will be generated based on road networks, terrain and population centers in Colorado, but shape files could be used from any region of the world. The results will generate various datasets that can be used for training analysts on IED attacks specific to any Operating Environment. This simulation will expose analysts to realistic data better preparing them for combat operations in their theater.