Performance of Statistical Power Control With Simplified User Selection and Mode Switching in Hybrid NOMA/OMA Networks

The paucity of orthogonal resource blocks has led to focused research on non-orthogonal multiple access (NOMA) signaling. This paper considers a multi-user downlink hybrid non-orthogonal multiple access (NOMA) framework with direct links to near-users (NUs) as well as far-users (FUs). It is shown that intelligent power allocation, together with mode switching and user selection, is the key to high spectral efficiency (SE) and energy efficiency (EE). To enhance the system throughput, intelligent NU and FU selection, and switching among NOMA Mode (NM), cooperative NOMA mode (CNM), and OMA mode (OM) have been considered. Switching to OM when both NM and CNM fail ensures performance equivalent to OM. The closed-form expressions for the outage probability, mode probability, throughput, and EE are derived with mode switching and user selection. It is demonstrated that choosing the optimal target rate and NOMA power allocation coefficient is essential to maximize the EE. Only statistical channel knowledge is used for this purpose. Furthermore, we demonstrate that the proposed framework outperforms known schemes in terms of throughput and EE. In particular, the energy savings with the proposed scheme can be very significant. Monte Carlo simulations validate the derived expressions.