Paper Title: Agent-Based Modeling for Testing and Designing Novel Decentralized Command and Control System Paradigms

Agent-based modeling (ABM) is a recent simulation modeling technique that consists of modeling a system from the bottom up, capturing the interactions taking place between the system’s constituent units. Such a bottom up approach enables users to describe and predict emergent phenomena. These include aggregate, system-level behaviors that can be counter-intuitive. Because decentralized command and control (DC2) paradigms can sometimes lead to counter-intuitive phenomena, ABM is the tool of choice to test DC2 and can provide significant insight into the design of DC2 approaches. 1. Decentralized Command and Control: Benefits and Challenges Testing new decentralized command and control (C2) paradigms can be challenging as decentralization requires a fresh mindset. New C2 methods can lead to unexpected, unanticipated system-wide behavior – also called emergent phenomena. While decentralization and distributed organizations offer clear benefits over traditional C2 approaches (robustness, flexibility, fluidity, responsiveness, self-organization), they also present clear challenges when it comes to designing the behavioral rules that, for example, individual soldiers must follow on the battlefield. For example, how does a commander test the rules of engagement to make sure that the system (ranging from platoon to brigade, or from swarms of unmanned aerial vehicles (UAV) to fleets of warships) can achieve its mission, meet specific requirements and not break down under rare but not impossible pathological conditions? Only when reliable testing tools become available will decentralized C2 paradigms be adopted. Such tools are necessary to reach the required level of confidence in new approaches and shift from a centralized C2 mindset to a decentralized one. An important solution to this dilemma has surfaced in the last few years: agent-based modeling (ABM) (Bonabeau, 2000, 2002a, 2002b; Casti, 1997; Epstein & Axtell, 1996; Hunt, 2001). In ABM, systems are modeled as collections of autonomous decisionmaking entities, called agents. Each agent individually assesses its situation and makes decisions based upon a set of rules. Agents may execute various behaviors appropriate for the system they represent. These systems, for example, may include searching, attacking, or performing battle damage assessment for an agent-based model of sensors such as UAVs. Repetitive, competitive and cooperative interactions between agents are a feature of agent-based modeling, where such modeling relies on the power of computers to explore dynamics out of the reach of pure mathematical methods. At the simplest level, an agent-based model consists of a system of agents and the relationships between them. Even a simple agent-based model can exhibit complex behavior patterns and provide valuable insights about the dynamics of the real-world system that it emulates. In addition, agents may be capable of evolving, thereby allowing unanticipated behaviors to emerge. Sophisticated ABM sometimes incorporates neural networks and genetic algorithms to allow realistic learning and adaptation. In recent years, ABM has been successfully applied to a range of problems in the commercial and civilian world (supply chain modeling, crowd modeling for evacuation, flow management in public spaces, theme parks and supermarkets, traffic control, organizational modeling, risk management, and market modeling) as well as in the military and law enforcement world (battlefield simulation, UAV control, military communications network security, and drug trafficking, for example). In the rest of this paper we will present the fundamental principles of ABM, present some applications of ABM relevant to command and control, and introduce ways of designing new DC2 paradigms using ABM as a testing platform. 2. Emergent Phenomena and Agent-based Modeling Emergent phenomena result from the interactions of individual entities. By definition, they cannot be reduced to the system’s parts: the whole is more than the sum of its parts because of interactions among the parts. An emergent phenomenon can have properties that are decoupled from the properties of the part. For example, a traffic jam, which results from the behavior of, and interactions between individual vehicle drivers, may be moving in the direction opposite to that of the cars that cause it. This characteristic of emergent phenomena makes them difficult to understand and predict: emergent phenomena can be counterintuitive. We will review a variety of examples of counterintuitive emergent phenomena in the following sections. Agent-based simulation is by its very nature the canonical approach to modeling emergent phenomena: in an agent-based simulation, one models and simulates the behavior of the system’s constituent units (the agents) and their interactions, capturing emergence when the simulation is run. Perhaps the simplest illustration of emergence is Boids, Craig Reynolds’ virtual creatures (Reynolds, 1987). Boids move in flocks (think of a flock of boids as a simple model of a flock of birds or a school of fish), they obey four simple rules, which characterize their behavior entirely: • Separation: steer to avoid crowding local flockmates • Alignment: steer towards the average heading of local flockmates • Cohesion: steer to move toward the average position of local flock-mates • Avoidance: steer to move away from oncoming obstacles When simulated in a computer, these four simple rules lead to an appearance of global coordinated behavior strongly reminiscent of real flocks or schools. For example, boids collectively avoid obstacles, a property that is not explicitly programmed into their behavior. In other words, collective obstacle avoidance is an emergent property.