Service-Context Knowledge-Based Solution for Autonomic Adaptation

In most adaptive systems, the adaptation control is based on developer-made rules and strategies that are specific for each service and context. Our proposal for autonomic computing is to replace this mechanism with a machine-based reasoning. The key element in making this possible is a service-context model that offers a knowledge support for the adaptive platform, which can diagnose the service adequacy to the context and search for solutions. We have tested our model using a prototype that adapts a service by inserting the 'right' component at the 'right' place into the service architecture.