The Department of Homeland Security has a strong interest in utilizing innovative simulations for planning the distribution of resources in order to prevent and defeat attacks from adversarial actions.  In this paper a creative Attack-Defense Tree (ADTree) capability is presented.  It allows an analyst to define any adversarial/defending scenario and conduct probabilistic computations for determining the effectiveness of the defensive structure.  A dedicated graphical user interface (GUI) makes it easy to define adversarial attackers, their attributes and the associated capability levels, their coordinated actions, their objectives, and all alternative paths for reaching their objectives.  The defensive structure is also defined in the GUI.  It includes all defense elements that are deployed along the potential paths of the attackers, their attributes and the associated capability levels, and how the adversarial attributes of the attackers compete against the defense capabilities of each element.  Coordination of resources among defense elements and the use of sensors and other detection capabilities as part of the defense infrastructure is defined in the GUI.  The ADTree conducts the probabilistic computations while considering all possible ways that each threat can accomplish its objectives and computes the probability of success of the threats.  The match-up between the capabilities and strengths of the adversaries’ vs the defense is considered in the computations at every step along each path that leads an adversary to an objective.  The detection of a threat is also accounted in the analysis.  The innovative theoretical elements of the probabilistic computations that provide flexibility for modeling any scenario of interest are presented and several examples are discussed.  The latter include an active shooter at a school, a cybersecurity example, and a border patrol case.  The examples present the utility of the ADTree in determining the most efficient approaches for hardening a defense.

Keywords
ASSESSMENT,DECISION,PROBABILITY,SECURITY,SIMULATIONS,VULNERABILITY MODELS

Additional Keywords