UAV-Enabled Non-Orthogonal Multiple Access Networks for Ground-Air-Ground Communications

Both unmanned aerial vehicle (UAV) and non-orthogonal multiple access (NOMA) have gradually become promising technologies for the fifth generation (5G) driven green Internet-of-Things (IoT) networks on account of their unique advantages of massive connections, higher spectral efficiency and flexibility. Motivated by this, we propose a 3-hop NOMA UAV-aided green communication network framework, where UAVs serve as aerial relays to support two groups of ground users. A stochastic geometry approach is invoked to model the spatial positions of the two group users. Under the realistic assumption, imperfect successive interference cancelation (ipSIC) is considered. To evaluate the performance of the proposed framework, theoretical expressions are derived to facilitate the outage performance evaluation of the far user (FU) and the near user (NU). Moreover, the asymptotic behaviors for the outage probability (OP) of both the FU and the NU in the high signal-to-noise ratio (SNR) regime are explored by obtaining diversity orders. Finally, the system throughputs under the delay-limited transmission mode are investigated. Numerical results confirm that: 1) For uplink transmission, there exist outage floors for the OP of both ipSIC and perfect SIC (pSIC) due to interference from the NU; 2) For downlink transmission, an outage floor exists for the OP of the NU under the condition of ipSIC; 3) For uplink NOMA/orthogonal multiple access (OMA) transmission, the outage performances of both the FU and the NU with NOMA outperform OMA in the low SNRs, while OMA has better performance in the high SNR regime; 4) For downlink NOMA/OMA, the outage performances of both the FU and the NU under pSIC outperform OMA.