An evolving asymmetric game for modeling interdictor-smuggler problems

Abstract

We propose a novel network interdiction model that reconciles many operational realities identified by military literature. Specifically, we conduct network interdiction within a dynamic network under partial information, using incomplete feedback and allowing two-sided adaptive play. Combining these aspects in an evolving game, we use optimization, simulation, and stochastic models to achieve a hybrid model. Modeling some currently underrepresented martial problems in this way makes it possible to highlight otherwise obscure relationships between policy and outcome, and to discover emergent effects, such as deterrence. As an example of this class of problems, we consider the struggle between a smuggler and interdictor. The smuggler seeks to maximize the amount of forces and materiel infiltrated from an origin to destination. The interdictor seeks to minimize this smuggler flow. Using two simple examples of an illicit-trafficking network, we demonstrate how to use these quantitative models within such an interdictor-smuggler context to (1) evaluate the value of seizures as a proxy for smuggled materiel, (2) assess the value of exploration, and (3) provide decision makers with practical ways to better allocate resources and increase effectiveness.