LAUSR

In this paper, we analyze the performance of a non-orthogonal multiple access (NOMA) cognitive relay system, where a multi-antenna full-duplex cognitive transmitter employs NOMA concept to assist the transmission from a wireless-powered primary transmitter to its corresponding receiver, while simultaneously communicating with a cognitive receiver. A practical non-linear energy harvesting (EH) model, which taking into account harvester's sensitivity and saturation effects is considered. We propose an optimum beamforming design at the cognitive transmitter such that the rate of cognitive network is maximized, under a constraint that the rate of the primary network is above a certain threshold. Results show that proposed optimization framework can substantially enlarge the rate region toward both primary and secondary networks. Furthermore, in order to characterize the network delay-constrained throughput, tractable outage probability expressions for primary and secondary networks assuming sub-optimum zero-forcing based beamforming scheme are derived. Our results reveal that due to required minimum power for harvesting operation as well as saturation of the harvested power at high transmit power levels, conventional linear EH model may lead to performance mismatches for practical non-linear EH circuits in both low and high transmit power regimes.