Performance of Full-Duplex Cooperative NOMA Network With Direct Link and Battery-Assisted Non-Linear Energy Harvesting Near User

This study investigates the performance of cooperative relaying-based non-orthogonal multiple access (NOMA) with simultaneous wireless information and power transfer. Assuming the batteryassisted non-linear energy harvesting (NL-EH) in which the near user (NU) works as a relay for the far user (FU) to convey data from the base station (BS); two scenarios are examined: (1) No direct link (NDL) between the BS and the FU, and (2) direct link (DL) between the BS and the FU. The DL at the FU is evaluated in two scenarios: as an optimal combination to the signal from the NU and as an interfering link to the signal from the NU. Considering the power splitting relaying (PSR) protocol with an NL-EH model and assuming imperfect successive interference cancellation (i-SIC) at the NU, we derive closed-form expressions for the NU and FU outage probability and throughput. The throughput of the FU is greatly increased by supplementing the harvested energy with a small quantity of battery energy. Furthermore, an optimal combination of the FU signals relayed from the BS and the NU considerably improves the throughput of the FU. It reduces the energy consumption at the battery compared to the scenario when there is NDL. We obtain the diversity order by deriving high signal-to-noise ratio approximations for the NU and FU outage probability in NDL and DL scenarios. Unlike linear energy harvesting, we show NL-EH leads to a loss of diversity, making it even more important to use the DL. We have formulated an equation to measure energy efficiency using the derived throughput expressions for the NDL and DL scenarios. We have also shown that selecting the optimal combination of battery energy and PS parameter is essential to achieve the highest possible throughput and energy efficiency. The residual interference generated by i-SIC significantly impacts the outage probability of the NU while having a negligible effect on the performance of the FU. Moreover, we have discovered that performance deteriorates when interference from the DL is present. We explain how the system's performance is affected by the selection of energy harvesting parameter, power allocation coefficient, saturation threshold, self-loop interference, variable node distances and battery energy. Monte Carlo simulations have verified the accuracy of the resulting expressions.