Motivation

We illustrate here some of the network science-related technical challenges faced by the modern Army with a small hypothetical example drawn from counterinsurgency (COIN), and present some of the key research questions that directly arise from this example.

Example Mini-Scenario

Intelligence officer Joanna Argus is planning the invitation of key tribal leaders for an urgent meeting to promote an alliance to stabilize the region’s governance. To maximize the chances of success, it is essential that the meeting be kept small, happen quickly, and include the people whose decisions will positively influence the tribes involved. The selection of invitees, and the timing and method of their invitations, will affect not only who
attends, but also who is antagonized by not being invited or informed, and the opportunities insurgents may have to disrupt the meeting or counter its results. As always, her decision must be based on information that is, to varying degrees, obsolete, incomplete, and inaccurate.

Supporting her expert judgment is a wide suite of technical issues, which currently have no scientifically based, reliable resolutions. The interdisciplinary research of the NS CTA will directly and qualitatively improve the ability of our networked systems to provide answers and improve her insight and decisions despite the tactical constraints of time and resources. The fundamental technical issues must address the means of achieving the most effective use of the many interworking networks (social/cognitive, information, and communications) of the Army, coalition, and host country; they must also address achieving the most accurate feasible understanding of and influence on the intimately intertwined networks used by the tribes and by the insurgents. Key NS CTA research directions include:

  • How can we model complex behaviors of composite networks (such as the propagation of trust and of trusted information and influence) to allow accurate prediction? How can we best estimate the current actual parameters for such models? In this example, such models would allow Argus to identify who will learn of the meeting, when, under what scenarios. What is the risk that a critical person will not receive an invitation from someone they trust in time to attend? Will a sensitive leader indirectly learn of the meeting before receiving an invitation? What are the insurgent networks embedded in those of the larger community? Is there an achievable scenario in which insurgent leaders will not receive trusted information about the meeting’s purpose and attendees in time to disrupt it? These questions affect not only how Argus plans the invitation process, but also even who is invited.
  • How can we configure our available network resources (of all genres) to achieve maximal effectiveness? The effectiveness of their use must be defined in terms of the social roles of information sources, potential attendees, non-invited neutrals and allies, and the (multiple) insurgent networks. Critical theoretical properties of composite networks of networks, defined and studied in the NS CTA, include the inherently crossdiscipline
    concepts of Quality of Information (QoI) and operational information content capacity (OICC). Argus has limited communications and information capabilities at her disposal, and limited knowledge of the social, information, and communications networks of the tribes and the insurgents. What achievable configurations of networks will provide Argus with the most critical information in the available time? What
    configurations will minimize the risks and make best use of delays when she distributes invitations and when the attendees propagate the results of the meeting?
  • What is the transient effect of changes in the parameters of one network on the behavior of other networks and the behavior of their integrated network of networks? Argus must act quickly, and must anticipate and influence the immediate consequences of her actions: if communications resources are redirected to provide better OICC to particular players, how will their social roles and relationships react? Do we risk destabilizing the social influence of a well-inclined tribal leader by devoting more communications or information resources to others in pursuit of better intelligence or expanded attendance?
  • How can we predict and influence the co-evolution of multiple genres of networks, each driven by different goals and capabilities? For Argus, such a capability could provide substantive answers (with bounded uncertainties) to such questions as: How will social results of the meeting (such as increased collaboration between allied tribes) drive the evolution of their intertwined multi-genre networks? How will the composite insurgent networks evolve in response to the changing configurations and uses of our networks and others? When insurgents learn of the invitations, how may they modify their communications, information, and social networks to disrupt or neutralize the meeting? How can Argus shape the short-term use of networks (in pursuit of this meeting) to promote longer-term co-evolution in favorable directions?