Energy harvesting cooperative diamond channel

We consider the energy harvesting diamond channel, where the source and two relays harvest energy from nature. The physical layer is modeled as a concatenation of a broadcast and a multiple access channel. We find the optimal offline transmit power and rate allocations that maximize the end-to-end throughput. First, we show that there exists an optimal source power allocation which is equal to the single-user optimal power allocation for the source energy arrivals and does not depend on the relay energy arrivals. Second, we show that the fraction of the power spent on each broadcast link depends on the energy arrivals for the relays. Specifically, we show that the optimal source rate allocation can be found by solving an optimal broadcasting problem with slot-dependent user priorities and these priorities can change only at instants where one of the relay data buffers is empty. Finally, we decompose the problem into inner and outer optimization problems and solve the overall problem by iterating between the two.