Stochastic transmission policies for energy harvesting wireless nodes

We consider a single-link wireless communication system where the transmitter has a finite-capacity rechargeable battery that is receiving energy packets from an energy harvesting device (EHD) at certain time instants. The transmitter may be a sensor node, which is responsible for sensing certain inputs from the environment and send them to the base over an additive white Gaussian noise (AWGN) or a fading channel. We pick the system throughput until a transmission deadline as the utility function, since higher throughput yields a higher service quality in general. A random energy leakage is assumed during the energy transfer from the EHD to the energy buffer of the transmitter. For cases where the energy leakage can be described by statistical distribution models, we formulate the parametric average throughput maximization problem with probabilistic energy causality and battery capacity constraints. We present problem feasibility conditions which are affected by the introduced threshold parameters. Furthermore, we propose an iterative algorithm to solve P and develop parameter dependent stochastic transmission policies (STPs) to manage the transmission power over a finite transmission interval. Simulation results indicate that STP improves the collective throughput of numerous single link energy harvesting nodes, operating over both the AWGN and fading channels compared to other leakage unaware power allocation techniques.