Predictive Power Management for Internet of Battery-Less Things

Abstract

Energy harvesting technology provides a promising solution to enable internet of battery-less things (IoBT), as the lifetime and size of batteries become major limiting factors in the design and effective operation of internet of things (IoT). However, with constrained energy buffer size, the variation of ambient energy availability and wireless communication cast adverse effect on the operation of IoBT. There is a pressing demand for developing IoBT-specialized power management. In this paper, we propose a novel predictive power management (PPM) framework combining optimal working point, deviation aware predictive energy allocation, and energy efficient transmission power control. The optimal working point guarantees minimum power loss of IoBT systems. By predictively budgeting the available energy and using the optimal working point as a set-point, PPM mitigates the prediction error so that both power failure time and system power loss is minimized. The transmission power control module of PPM improves energy efficiency by dynamically selecting optimal transmission power level with minimum energy consumption. Real-world harvesting profiles are tested to validate the effectiveness of PPM. The results indicate that compared with the previous predictive power managers, PPM incurs up to $17.49\%$ reduction in system power loss and $93.88\%$ less power failure time while maintaining a high energy utilization rate. PPM also achieves $9.4\%$ to $23.22\%$ of maximum improvement of transmission energy efficiency compared with the state-of-the-art transmission power control schemes.