A Comparison between Two Approaches to Threat Evaluation in an Air Defense Scenario Fredrik Johansson 1 and G¨ oran Falkman 2 1 University of Sk¨ovde fredrik.johansson@his.se 2 University of Sk¨ovde goran.falkman@his.se Abstract. Threat evaluation is a high-level information fusion prob- lem of high importance within the military domain. This task is the foundation for weapons allocation, where assignment of blue force (own) weapon systems to red force (enemy) targets is performed. In this paper, we compare two fundamentally different approaches to threat evaluation: Bayesian networks and fuzzy inference rules. We conclude that there are pros and cons with both types of approaches, and that a hybrid of the two approaches seems both promising and viable for future research. Keywords: Bayesian networks, fuzzy inference rules, fuzzy logic, threat assessment, threat evaluation, weapons allocation. 1 Introduction In a military environment it is often the case that decision makers in real-time have to evaluate the tactical situation and to protect defended assets against enemy targets by assigning available weapon systems to them [1]. In situations with several potential threats, a prioritizing of targets is necessary. Such an order of priority is often based on the degree of threat the targets represent to friendly defended assets. To determine which of several threats that represent the highest danger is of great importance, since errors such as prioritizing a lesser threat as a greater threat can result in engaging the wrong target, which often will have severe consequences [2]. The calculation of such threat values is often referred to as threat evaluation. Threat evaluation is a part of threat analysis [3], which in an information fusion context is a central part of level 3 (impact assessment) in the well-known Joint Directories of Laboratories data fusion model [4]. Threat evaluation is the basis for weapons allocation, a process in which the decision-maker decides on which weapon system that should be assigned to a certain target. Research in high-level information fusion is still relatively immature [5]. As a consequence, different methods have been proposed for e.g. threat evaluation, but a systematic comparison between different approaches is lacking. Therefore, in this paper, we implement and compare two different artificial intelligence (AI) approaches to threat evaluation. The first method is the Bayesian network approach described V. Torra and Y. Narukawa (Eds.): MDAI 2008, LNAI 5285, pp. 110–121, 2008. c  Springer-Verlag Berlin Heidelberg 2008