The impact of finite data and energy storages in performance of relay networks with energy harvesting

We consider a wireless network consisting of a source, an energy harvesting relay and a destination. A discrete time schedule is assumed. In the first stage, the energy packets and the data packets are sent to the relay node and stored in the finite battery and data storages, respectively. In the second stage, the stored data is sent towards the destination. It is supposed that at each time slot, only a packet of energy or data could be transmitted. The probability of successful transmission is derived for both data and energy packets. An acknowledgment packet reports the status of the buffers to the source node, based on the probability of successful transmissions. The network Packet Delivery Ratio (PDR) is considered to evaluate the performance of the system. Thus, a maximization problem is defined over discrete parameters of the network on data and energy buffer constraints. The optimization problem turns out to be non-convex, thus, it is converted to the subproblems which are solved individually. The closed form expressions are derived to determine the number of the packets during each time frame. The analytical results are verified by the simulations and show the effect of energy and data capacity on PDR.