QoS-Driven Power Control for Energy Harvesting Fading Multiple-Access Channels

This paper investigates the energy harvesting fading multiple-access channels, where the transmitters harvest random amounts of energy from the environment and the data transmissions are subject to statistical quality of service (QoS) constraints in the form of limitations on the buffer overflow probability. With the assumption that the channel state information (CSI) and harvested energy in all time slots are known at the transmitters, the point-to-point link is first revisited to obtain the effective capacity expression with the optimal offline power control policy. Then, regarding the transmissions with super-position coding, the conditions that the offline optimal power control policies must satisfy for a given decoding order strategy are determined. A suboptimal online power control policy is proposed. It is shown that the proposed power control policies can achieve performance close to the optimal one achieved with average power constraints only in case of infinite energy buffer.