Abstract

Computational frameworks motivated by the efficiency of biological systems abound---using mechanisms adapted from models of naturally occurring behavior such as foraging, immuno-response, dominance contests, behavioral thresholds for task performance, worker recruitment, etc. Often, the motivation is as simple as that it works well for the biological system's problem and perhaps a similar mechanism can work just as well for a problem-solving framework inspired by the biology. These nature-inspired systems are often robust, effective problem solvers. In this paper, we consider the application of game theory for the analysis of biologically-inspired agent coordination mechanisms. The objective of this paper is to provide examples of how game theory can be used to explain the emergent behavior of collective problem solving systems. To illustrate, we analyze an existing multi-agent task allocation protocol motivated by a computational model of wasp behavior.