Downlink base station cooperation with energy harvesting

This paper considers a downlink cellular network with two base stations (BSs) powered by harvested energy to cooperatively serve two users. To tackle the problem that BSs are of different energy arrival rates, each transmission frame is divided into two subframes, i.e., in the first subframe, one of the BSs keeps silent for some time to store energy while the other transmits to one of the users, and then they perform zero-forcing joint processing (ZF-JP) in the second subframe. We propose an algorithm to optimize the sum rate by jointly deciding which BS should store the energy, how long does it keep silent, and how much transmit power should be used in single BS transmission mode and ZF-JP mode, respectively. Simulation results show that the proposed algorithm can achieve much higher sum rate compared with the ZF-JP only scheme, and the gain increases as the difference between BSs´ energy arrival rates grows.