FESAS: A Framework for Engineering Self-Adaptive Systems
Christian Krupitzer · MADOC (University of Mannheim) · 2013
Abstract—The complexity and size of information systems are growing, resulting in an increasing effort for maintenance. Autonomic Computing, Organic Computing or self-adaptive systems (SAS) that autonomously adapt to environment changes or changes in the system itself (e.g. disfunction of components) can be a solution. So far, the development of a SAS is tailored to its specific use case requirements. The creation of frameworks with reusable process elements and system components is often neglected. However, with such a framework developing SAS would become faster and less error prone. This work addresses this gap by providing a framework for engineering SAS. The framework is based on model-driven engineering and a service-oriented architecture. At the development stage, a design model is transformed into a system model. During runtime, this system model is mapped to a concrete system build of various heterogeneous devices with functionalities. The functionalities of the devices is represented as services. A reference architecture and a component library support this mapping and enable the deployment of an adaptation logic tailored to the system’s use case. This reduces engineering time through a generic process with reusable elements.