Multiagent Cooperation in International Coalitions

increases confusion in the opponent’s mind. This process is command-led, with human decision making primary and technology playing a secondary role. Shared understanding and information superiority are key enablers in this process and are fundamental to network-centric warfare (www.dodccrp.org). In addressing interoperability requirements, we must also address data security, control over semitrusted software from other coalition partners, and the resulting system’s robustness—for example, its ability to withstand denial-of-service attacks. However, good decisions depend on good data, which can become a problem as mission complexity increases. Military coalitions—large-scale, multifaceted, multinational, virtual organizations—often must be rapidly created and changed as circumstances shift. In addition to integrating single-service and joint capabilities into a coherent force, coalition operations must rapidly configure foreign or legacy systems into a cohesive whole. Yet current coalition operations often suffer from data overload, information starvation, laborintensive information collection and coordination, and stand-alone stovepipe command systems that use incompatible data formats. This leads to a horrendous technical integration task and offers commanders only scattered snapshots of the battlespace. In the inevitable absence of preexisting coordinated systems, we must take a rapid, flexible, on-the-fly approach that permits assembling capabilities at runtime. We believe agent-based computing (described in the sidebar, “Software Agent Technology”) offers a promising new approach to effective coalition operations because it embraces the coalition environment’s open, heterogeneous, diverse, dispersed nature. This article describes how software agents acting on behalf of human users enable military commanders to act decisively in cyberspace and thus contribute to “cyberspace superiority,” a critical component of warfare in the information age.1 We focus here on the rapid integration of agents and legacy systems to improve interoperability and support human situational awareness and decision making, rather than, for example, sophisticated teamwork and planning between agents.2