A cognitive agent-based approach to varying behaviours in Computer Generated Forces systems to model scenarios like coalitions

In this paper we describe a general approach for dynamic variation in behaviour of entities in military-based Computer Generated Forces (CGF) system scenarios. A typical scenario might represent the coalition of military forces from two countries to provide suppression of an enemy's air defence capability. The entities in the scenario are cognitively-plausible models of individual people and hierarchical groups of people, and are modelled using intelligent software agents. The cognitive agents' Situational Awareness (SA) of the ongoing battle is affected due to the imposition of various moderating factors such as fatigue, fear and stress. This variation of SA-impinges on the agents' selection and execution of tactics in response to the observation of threats and so provides the CGF system with a higher degree of realism. This work is central to the development of the next generation of Computer Generated Forces systems that are to be inherently adaptable, 'intelligent' and dynamically re-configurable in order to cope with the fast changing demands of military Operational Analysis, procurement and training environments. The paper discusses the key features of these cognitive agents' variability through a military demonstration.