Enabling Efficient Communications with Resource Constrained Information Endpoints in Smart Homes

Smart Homes are one of the most promising real applications of Internet of Things and Cyber-Physical Systems. Devices and software components are deployed to create enhanced living environments where physical information is captured by sensors, sent to servers and finally transmitted to information endpoints to be consumed after its processing. These systems usually employ resource constrained components in dense architectures supported by massive machine type communications. Components, to adapt to different scenarios, present several configuration options. In machine type communications, these configuration options should be selected dynamically and automatically. Many works have addressed this situation in relation to sensor-server communications but endpoints are still mostly manually configured. Therefore, in this paper it is proposed an automatic and dynamic configuration algorithm, based on the idea of "efficiency," for information endpoints in the context of Smart Homes. Different costs associated to endpoint-server communications in Smart Homes are identified and mathematically modelled. Using this model and real measurements, the most efficient configuration is selected for each endpoint at each moment, not only guarantying the interoperability of devices but also ensuring an adequate resource usage, for example, modifying the endpoints' lifecycle or the information compression mechanism. In order to validate the proposed solution, an experimental validation including both real implementation and simulation scenarios is provided.