Lightweight operation scheduling for self-powered IoT devices

Abstract

Perpetual operation is a potential necessity in self-powered IoT devices. System level low power optimization can significantly decrease the power consumption and increase system lifetime. Further, energy scavenging technologies have been proposed to improve the lifetime of such a system, up to perpetual operation. However, fixed duty cycle (active/sleep) operation can cause inefficiencies in the presence of variable input power; further, in cases of limited storage capacity and intermittent input power, the system’s perpetual operation is not guaranteed. Energy forecasting algorithms can propose adaptable duty cycles, at the expense of heavy computing workloads. In this paper, we propose adaptable duty cycle operation for energy efficient IoT sensor nodes, utilizing a very simple -and very lightweight - scheduling algorithm that depends only on the periodical provision of the accumulator’s state-of-charge. The proposed algorithm achieves energy balance in the presence of harvested energy and guarantees a minimum operation time in the absence of harvested energy.