Chapter 11 - Knowledge Representation, Processing, and Governance in the FOCALE Autonomic Architecture

Abstract

This chapter describes the design and evolution of the FOCALE autonomic architecture. FOCALE has been designed for managing heterogeneous networks. However, its principles can be used for other applications, as the mechanisms described are not specifically oriented toward network management. In particular, the most difficult challenge that FOCALE has overcome—integrating heterogeneous management and operational data from different sources that use different languages to describe their data—makes FOCALE well suited to being used in other industries, since the integration of disparate management and operational data was done using a novel new knowledge representation that was independent of any underlying technology being managed. This chapter emphasizes the importance of knowledge representation and fusion. The FOCALE approach recommends integrating models and ontologies. Models provide a powerful way to describe facts and support advanced code-generation methodologies. Ontologies represent a powerful set of tools to attach semantic meaning to data. The FOCALE architecture is a specific architecture that uses models and ontologies to represent knowledge. The Model-Based Translation module converts vendor-specific sensor data into a common vendor-neutral form, and vendor-independent commands into vendor-specific versions. The autonomic manager determines the state of the entity being managed from received sensor data and/or inferences, and determines what actions to take guided by context-aware policy rules. Semantic reasoning helps orchestrate system behavior through the use of machine-based learning and reasoning.