An approach for knowledge-based IT management of air traffic control systems

In this paper, an approach for runtime analysis and automated knowledge-based IT management is presented and applied to the example of an Air Traffic Control (ATC) system. An analysis method has been developed, which combines the strengths of formal ontologies and Complex Event Processing (CEP). Thereby, the configuration of the whole analysis process is derived from one homogeneous knowledge source, a formal ontology model, which contains the system topology, the events that occur in the system, aggregations and Service Level Agreements (SLAs). Ontology rules are evaluated using semantic reasoning, event aggregations are translated into rules for the CEP engine. During runtime, data from the Managed System is classified to be either low-frequent or high-frequent and is accordingly either mapped to the ontology or used as input for the CEP engine. The aggregated information from both processes is used as input for the planning of reconfigurations on the Managed System.