Optimal Harvest-Use-Store Strategy for Energy Harvesting Wireless Systems

Energy harvesting (EH) technology has emerged as a promising energy-supplier to unattended wireless systems. In the commonly used harvest-store-use (HSU) scheme, harvested energy is always stored in a battery before its subsequent use. The existence of storage loss in practical battery systems, however, unavoidably reduces the energy efficiency. In this paper, we therefore propose the use of a more efficient harvest-use-store (HUS) architecture for point-to-point data transmission, where the harvested energy is prioritized for use in data transmission while its balance/debt is stored in or extracted from the storage device. We derive the optimal energy polices, under the criterion of throughput maximization, for the HUS architecture on static and block fading channels, and investigate the properties of the resulting power allocation pattern. The optimization is done in the Lagrangian framework, uncovering the special structure of the optimal power pattern and obtaining a closed-form solution conditioned on the knowledge of the block locations for zero battery level. A dynamic programming (DP) based algorithm is developed for locating such blocks in the optimal power patterns. Numerical results are presented to demonstrate the properties of the proposed HUS architecture and its superior performance over the existing schemes.