Fading channels in energy-harvesting receivers

We consider the operation of an energy harvesting receiver in a point-to-point fading channel with additive white Gaussian noise (AWGN). Knowledge of the channel state information is not available at the transmitter. The limited rate of harvesting energy at the receiver along with the time variation of the channel can degrade the performance of the system. However, we will show that channel state knowledge at the receiver can improve the performance of the system. We propose a channel-selective sampling strategy that optimizes a tradeoff between the energy costs of sampling and decoding at the receiver. Based on this tradeoff, we derive a policy maximizing the communication rate and we characterize an energy-constrained rate region. We extend the results to the receivers with finite battery capacity.