A Framework for Executing Computational Intelligence Over Distributed Embedded Nodes
Robert Ellen, Duncan Campbell · QUT ePrints (Queensland University of Technology) · 2005
Manufacturing and mining automation, robotics, swarms and smart device networks are often implemented upon distributed embedded systems. These systems are typically statically distributed, coarsely recongurable or deployed on homogeneous networks. A conceptual stack can be formed using modelling languages, system performance analysis and optimisation and recongurable platform-neutral components to overcome these problems. A framework for computational intelligence-based applications to be built upon this stack has been proposed. Model-Driven Architecuture has been shown to be a promising standard for the modelling, design and development of embedded applications. The Theory of Constraints is proposed as a potential technique for performance analysis and optimisation of distributed systems. Mobile Agents and code mobility can be used in the component architecture to allow for adaptation and reconguration for optimisation.