Power splitting versus time switching based cooperative relaying protocols for SWIPT in NOMA systems

Abstract Non-orthogonal multiple access (NOMA) along with wireless power transfer have recently been adapted to cooperative communications for 5G and beyond wireless networks. This paper investigates NOMA based cooperative relaying wireless-powered networks (CRWPNs) where, decode-and-forward (DF) relaying and successive interference cancellation are both employed at a wireless-powered intermediate node. For simultaneous wireless information and power transfer (SWIPT), power-splitting relaying (PSR) and time switching-based relaying (TSR) protocols are considered in the NOMA based CRWPN. As a result, the combination of cooperative relaying power domain NOMA network and PSR and TSR protocols is proposed in this paper. The outage performance and ergodic rate of both protocols are analysed for evaluation of the impacts of energy harvesting (EH) time, EH efficiency, power splitting ratio, source data rate, and the distance between the nodes. In addition, two delay limited transmission (DLT) and delay tolerant transmission (DTT) modes are considered in this network model to investigate the throughput and ergodic rate of the system according to the source transmission rate. It is shown that the cooperative relaying NOMA (CRNOMA) scheme achieves a lower outage probability when compared to the conventional orthogonal multiple access (OMA) schemes. Additionally, the PSR outperforms the TSR in both low and high signal-to-noise ratio ( S N R ) regions in terms of throughput, ergodic rate and energy efficiency. For instance, the outage probability of CRNOMA for both PSR and TSR in S N R range of from -10 dB to +20 dB (i.e. a low S N R region) decreases gradually but not linearly. However, in S N R range of from +20 dB to +40 dB (i.e. a high S N R region), the outage probability of CRNOMA for both PSR and TSR decreases quickly. Furthermore, the energy efficiency is shown to be considerably enhanced with the employment of EH for CRNOMA. Finally, the impacts of the distance between the nodes on the performance and a comparison between two scenarios of having and without having direct links are evaluated.