A survey of multi-agent organizational paradigms
The Knowledge Engineering Review
Quantitative organizational models for large-scale agent systems
MMAS'04 Proceedings of the First international conference on Massively Multi-Agent Systems
| Hi-index | 0.00 |
We describe how a system employing different types of organizational techniques addresses the challenges posed by a large-scale distributed sensor network environment. The high-level multi-agent architecture of real-world system is given in detail, and empirical and analytic results are provided showing the various effects that organizational characteristics have on the system's performance. We show how partitioning of the environment can lead to better locality and more constrained communication, as well as disproportionate load on individuals or increased load on the population as a whole. The presence of such tradeoffs motivates the need for a better understanding of organizational effects.