Design Pattern for Self-Organizing Emergent Systems Based on Digital Infochemicals

Holger Kasinger, Bernhard Bauer, and Jörg Denzinger

appeared in: Proceedings Sixth IEEE Conference and Workshops on Engineering of Autonomic and Autonomous Systems, San Francisco, 2009, pp. 45-55.



Abstract

An essential element in the engineering of computer systems are design patterns that capture current best practice and knowledge about recurring solutions for standard problems. In case of decentralized autonomic computing systems, also known as self-organizing emergent systems, appropriate design patterns have to structurally describe decentralized coordination mechanisms along with information on which kind of macroscopic effects, the self-* properties, can be achieved in which situations. In this paper we present a design pattern for self-organizing emergent systems coordinating by means of digital infochemicals. Infochemicals, in the natural context, are chemical substances that convey information in the interaction between two individuals. Because infochemical coordination is the most universally employed mechanism of communication in nature, there exists plenty of inspiring examples of decentralized coordination usable for the solution of complex problems in need of certain self-* properties. The presented design pattern captures the general biological principles behind infochemical coordination, which simplifies a systematical systems engineering. It extends existing coordination models, in particular pheromone-based coordination and digital semiochemical coordination, in terms of terminology, functionality, as well as generality, and thus becomes applicable to a much wider set of problem domains.



Download paper (publisher's webpage)

Generated: 13/7/2009