Discerning Properties of a Self-Organizing Network (Swarm) Shaping its Structure, Function, and Resilience
Abstract
Advancing the development of a net-centric Army depends on our ability to understand the benefits and costs of information flow through networks. Networks can vary in composition and expanse, with a variety of examples present in daily life (e.g., the internet, acquaintances, animal groups). The functionality of a given network generally depends upon the connectivity among its constituent members (nodes). As with acquaintances and popular web sites, these connections often change over time. Most network analyses focus on static systems using snapshots of data taken in time. Investigators then search for correlations between the degree of connectivity within the network and the network's ability to resist interference (i.e., noise) or external perturbation (e.g., power loss, attack). We seek insight into the capability of network analysis to discern properties of a dynamic network (of collective behavior) shaping its structure, function, and resilience using swarming behavior as our model system. Specifically, in this paper we begin answering the question: how many neighbors should an individual in a networked swarm track for maximum efficiency of information transfer, critical to survival under attack and failure scenarios. Through numerical systems based on ecological theory we seek to better understand networked collective behavior, e.g., susceptibility to outside attack, which can be used to guide the engineering design of artificial networks. We believe this work presents an opportunity to mine the assets of ecology for improved development of disruptive technologies and a net-centric Army.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 2008
- Accession Number
- ADA505781
Entities
People
- B. Lemasson
- J. J. Anderson
- R. A. Goodwin
- T. S. Bridges
Organizations
- Portland District, U.S. Army Corps of Engineers