This paper examines the proposition that terrorist networks, such as Al Qaeda, are complex adaptive systems; that is, they consist of widely dispersed, autonomous cells that obey a decentralized command and control hierarchy; their mission operatives are highly adaptive and mobile; their cells are strongly compartmentalized, structurally robust, and largely impervious to (unfocused) local attack; and, though the networks, as a whole, are typically covert and amorphous, they can also rapidly coalesce into tightly organized local swarms.
This implies that, in principle, terrorist networks, as dynamical systems, ought to be amenable to the same methodological course of study as any other complex adaptive system (such as a natural ecology, a biological immune system, or the human brain). In particular, fundamental insights into the behavior of terrorist networks—including an understanding of how they form, how they evolve, how they adapt (to changing internal and external contexts), and what their innate strengths and vulnerabilities are—may be gleaned by studying the patterns that emerge from a multiagent-based simulation of their dynamics.
This paper has two primary goals: (1) to review existing analytical and modeling tools that are applicable to the study of dynamic networks (including mathematical graph theory, social network modeling, complex network theory, graph visualization, and multiagent-based modeling), and outline how these tools may be leveraged to help understand the dynamics of terrorist networks, and (2) to introduce the conceptual design of a new multiagent-based toolkit, called SOTCAC (Self-Organized Terrorist-Counterterrorist Adaptive Coevolutions). SOTCAC uses autonomous, intelligent agents to represent the components of coevolving terrorist and counterterrorist networks.
⇒ Graph Theory provides a mathematical formalism with which to represent arbitrary relationships among the components of a complex system; as well as a computational aid for discovering latent patterns embedded within those relationships.
⇒ Social Network Analysis studies patterns of relationships (such communication, information and resource flow, etc.) among individuals and organizations, and is particularly adept at revealing otherwise “hidden” patterns inside of a network; for example, information-flow bottlenecks and other vulnerabilities of, say, a business organization, that are not obvious from a wire-diagram of its members.
Finally, SOTCAC builds upon the multiagent-modeling technologies underlying both the EINSTein and SCUDHunt simulations, recently developed at the Center for Naval Analyses (CNA). Agent simulations provide a powerful, generative modeling environment for performing exploratory analyses on self-organized emergent behavior in complex adaptive systems.
~
ILACHINSKI, Andrew, 2005. Self-Organized Terrorist- Counterterrorist Adaptive Coevolutions, Part I: A Conceptual Design.