Open Access

The Internet of Things (IoT) has transformed already many areas of our modern life. IoT devices are being connected through IoT platforms and new applications & services are created based on IoT data to provide added value for customers. Interoperability between different IoT platforms is still one of the most important gaps inside the Internet of Things. Current interoperability solutions are still largely driven by static & inflexible interoperability mechanisms originating from the past webservice-era which do not fit the characteristics of modern IoT ecosystems. Such static, design-time approaches suffer from scalability issues since they are not suitable for large, distributed systems and their real-time constraints. A new, dynamic approach is required, which requires IoT system design to be reconsidered, enabling properties of autonomy and intelligence. This thesis addresses interoperability from such a dynamic perspective, based on Organic Computing concepts. An architecture for a self-adaptive IoT system will be presented which is able to maintain and improve runtime interoperability in an IoT ecosystem autonomously.