Summary: | The consolidation of the Fog Computing paradigm and the ever-increasing diffusion of Internet of Things (IoT) and smart objects are paving the way toward new integrated solutions to efficiently provide services via short-mid range wireless connectivity. Being the most of the nodes mobile, the node discovery process assumes a crucial role for service seekers and providers, especially in IoT-fog environments where most of the devices run on battery. This paper proposes an original model and a fog-driven architecture for efficient node discovery in IoT environments. Our novel architecture exploits the location awareness provided by the fog paradigm to significantly reduce the power drain of the default baseline IoT discovery process. To this purpose, we propose a deterministic and competitive adaptive strategy to dynamically adjust our energy-saving techniques by deciding when to switch BLE interfaces ON/OFF based on the expected frequency of node approaching. Finally, the paper presents a thorough performance assessment that confirms the applicability of the proposed solution in several different applications scenarios. This evaluation aims also to highlight the impact of the nodes' dynamic arrival on discovery process performance.
|