Description and composition of bio-inspired design patterns: the gradient case

Abstract

Bio-inspired mechanisms have been extensively used in the last decade for solving optimisation problems and for decentralised control of sensors, robots or nodes in P2P systems. Different attempts at describing some of these mechanisms have been proposed, some of them under the form of design patterns. However, there is not so far a clear catalogue of these mechanisms, described as patterns, showing the relations between the different patterns and identifying the precise boundaries of each mechanism. To ease engineering of artificial bio-inspired systems, this paper describes a group of bio-inspired mechanisms in terms of design patterns organised into different layers. This approach is exemplified through the description of 7 bio-inspired mechanisms: three basic ones (Spreading, Aggregation, and Evaporation), a mid-level one (Gradient) obtained by composing the basic ones, and three top-level ones (Chemotaxis, Morphogenesis, and Quorum sensing) exploiting the mid-level one.