Skip to Main Content
Renewable energy, such as solar radiation, can be used to extend the lifetime of a wireless sensor network (WSN) node. Such systems fundamentally change the problem of power scheduling. Instead of maximizing system lifetime with a fixed amount of energy as in battery-powered systems, the purpose of scheduling becomes the prevention of energy depletion at any given time. On the other hand, partial dynamic reconfiguration is demonstrated as an efficient technique for intensive applications, such as video and encryption processing. Unlike software (SW)-based implementation, reconfigurable hardware (HW) requires time and energy for reconfiguration before enabling the application, which causes a trade-off between the SW and the reconfigurable HW. Therefore, when reconfigurable HW is applied in an energy-harvesting system, the problem lies in the manner by which to schedule the dynamic reconfiguration, such that the dynamically harvested energy is efficiently used. In this paper, a novel methodology is presented for the scheduling of the dynamic reconfiguration for a WSN node under energy-harvesting conditions based on statistical information on tasks and available energy. To evaluate our method, an HW-reconfigurable WSN node is prototyped, and related experimental data are collected. Our experiments demonstrate that by implementing our method, more than 50% of the energy costs can be saved with a 50% performance improvement.