Emergence as a general architecture for distributed autonomic computing

Today’s systems are becoming more and more complex, i.e. distributed, situated, open, and dynamic. Autonomic computing aims to deal with the complexity autonomously. Hence, distributed autonomic computing systems tend to consist out of autonomous entities because of the increased distribution. This increased complexity and autonomy makes it difficult to build systems with a global coherent behaviour as a huge number of entities are expected to cooperate. There are signs of a future for emergence in multi-agent systems as a general architecture for the individual entities and the distributed autonomic computing system as a whole. Understanding and exploiting the process of emergence is key to study and engineer global coherent behaviour, and thus key to develop and manage distributed autonomic computing systems efficiently. As a consequence, there is a need for fundamental scientific methods to gain those insights and control tomorrows emergent autonomic computing systems. Chaotic time series analysis is one scientific method that can provide those fundamental insights.