Delay Constrained Energy Harvesting Networks with Limited Energy and Data Storage

This paper studies energy harvesting transmitters in the single user channel, the two-way channel, and the two-way relay channel with block fading. Each transmitter is equipped with a finite battery to store the harvested energy, and a finite buffer to store the data that arrive during the communication session. We consider delay sensitive applications and maximize throughput while enabling timely delivery of data with delay constraints. We show that the resulting delay limited throughput maximization problem can be solved using alternating maximization of two decoupled problems termed the energy scheduling problem and the data scheduling problem. We solve the energy scheduling problem using a modified directional waterfilling algorithm with right permeable taps, water pumps, and overflow bins and the data scheduling problem with forward induction. Additionally, we identify the online optimum policy for throughput maximization. We provide numerical results to verify our analytical findings and to demonstrate the impact of the finite data buffer capacity and the delay requirements on the throughput. We observe that larger buffer sizes become useful for more lenient delay requirements, and a data buffer size that is comparable to the throughput within one time slot accounts for the majority of the increase in throughput.