Reusable, semantic, and context-aware micro-architecture: Approach to managing interoperability and dynamics in smart spaces: Dissertation

The amount of shared information has increased a great deal in ubiquitous systems, where the previously isolated devices and appliances have become part of the system and are producing or consuming the information. The ubiquitous system, or the smart environment, lacks an approach that supports scalability and enables semantic interoperability. It is challenging to provide a dynamic behavior at the run time without human intervention. A number of dedicated solutions have been developed for the ubiquitous environment because of its complexity. The dedicated solutions are usually non reusable. An approach is needed that i) is reusable as such or partly, ii) provides the semantic interoperability, iii) enables dynamic and behavioral interoperability between the receiver and sender of the information at run time, and iv) is scalable by being modular, and decoupled. This thesis proposes a novel approach to managing interoperability and dynamics in smart spaces. The approach includes a Context-Aware Micro-Architecture (CAMA), and a Context Ontology for Smart Spaces (CO4SS). This approach is independent of implementation languages and communication techniques. CAMA, as an architectural pattern, is usable without its semantic support, CO4SS. In the literature, it is the first approach that fulfills the requirements that are set for a context data distribution system. The power in CAMA relies on the usage of the standard and web-based techniques, in the separation-of-concerns principle, and in the enhanced control loop, MAPE-K. The latter has four parts, Monitor; Analyze; Plan; Execute that share Knowledge. CAMA is highly dynamic, which is due to the run-time updatable rules. The creation of the rules is laborious, as they are written into text boxes of Message Sequence Charts. This will be improved when new tools are developed for the rule creation. Additional research is needed to validate the scalability of the approach with a "Big data". CO4SS can be widened with the domain-specific and quality ontologies. It supports the evolution management of the smart space: all smart spaces and their applications 'understand' the common language that is defined by it. CO4SS has the potential to be a de facto ontology for the context-aware, i.e., intelligent applications.