Leveraging Micro Energy Sources in Energy Harvesting Wireless Sensor Networks

Effectively utilizing very small energy sources, like ambient thermal gradients, in a Wireless Sensor Network (WSN) offers great potential for novel applications. Further, without over provisioning the energy source or storage of an Energy Harvesting-WSN, it can be operated more ecologic. We reduce the overhead associated with peripheral initialization by dynamically adapting the boot process of sensor nodes operating in a burst style operation from micro energy sources. For the evaluation, we implement a framework that allows reproduction of real-world conditions for Thermoelectric Generator energy harvesters. Based on real-world data our study shows an improvement of 27% in the number of tasks executed with only half the energy required for certain tasks. Finally, we discuss the factors influencing the effectiveness of our dynamic boot approach w.r.t application scenario.