On the Spectral Efficiency of Multiuser Scheduling in RF-Powered Uplink Cellular Networks

This paper characterizes the spectral efficiency of an uplink radio frequency (RF)-powered macrocell network considering harvest-then-transmit protocol such that the macrocell users transmit in the uplink while replenishing the energy from their serving base station (BS) in the downlink. Using the theory of order statistics, a tractable mathematical framework is developed to derive the uplink spectral efficiency and the downlink power consumption resulting due to wireless energy transfer. The framework captures the impact of the locations of the users that are selected for uplink transmission, their channel statistics for information and energy transfer, and different user selection schemes. We first analyze the performance of state-of-the-art greedy and round-robin scheduling schemes in RF-powered cellular networks. Closed-form expressions for the minimum power outage probability (i.e., the probability that the selected user is unable to harvest sufficient power for uplink transmission) are also derived. We then develop modified versions of the conventional user selection schemes that improve the spectral efficiency on a given uplink transmission channel with zero power outage probability (i.e., probability of outage due to insufficient amount of harvested power). The developed schemes are shown to outperform the conventional user scheduling schemes in terms of the throughput and energy harvesting time with a trade-off in fairness among users. The accuracy of the expressions is validated via Monte-Carlo simulations. Numerical results highlight the trade-offs associated with the various user selection schemes as a function of network parameters.