Communication Properties of Self-Organizing Networks (SWARMS) as Inferred from Optical Mechanics

摘要

A primary question in our research is to determine how much sensory processing is necessary to maintain a network's resiliency to fission, or collapse, while minimizing the degree of disorder in communication. Biological components capable of governing these processes include perception, thresholds, and simple decision filters. We present a self-propelled particle model (s.p.p.) with simplified optical rules in which individual interactions are dictated by retinal representations of object bearing and motion. Optical signals are weighted by individual perception and the degree of redundant neurological stimulation. In turn, signals that exceed a threshold value serve to identify influential neighbors within the swarm's global network. Individual connections (i.e., each influential neighbor) is normalized to the individual's immediate surroundings, making the decision process adaptive to both physical and behavioral variability.