Efficient auto-configuration of energy-related parameters in cloud-based IoT platforms

Cloud-based platforms for handling Internet-of-Things data can reduce the energy consumption of various ICT modules (e.g., data centers, networking equipment, battery-equipped devices) by applying appropriate system configurations. However, this is currently prevented by the higher prioritization of other goals, the complexity or unawareness of the synergies between system configurations and energy consumption, and the technical difficulties for automating re-configuration. This paper contributes a framework, a study of synergies between system parameters and energy-related goals, as well as an auto-configuration algorithm. It is the first solution that is based on the mentioned synergies and employs the concept of stylesheets for device-specific interpretation of system-wide configuration suggestions. The evaluation has shown that auto-configuration can be executed very fast and achieves much higher levels of fulfillment of operator targets than static configurations.