On the Power Allocation and Constellation Selection in Downlink NOMA

In non-orthogonal multiple access (NOMA), signals of different users are independently encoded and modulated, then superposed in the power domain before transmission. The choice of a suitable combination of power allocation factor (PAF) and transmit constellations is hence critical to enhance the system performance. The aim of this work is to develop a tool for adaptively selecting a suitable PAF and transmit constellations for an enhanced system data rate, given some individual user bit error rate (BER) constraints that need to be jointly considered. For this purpose, analytical bit error probability (BEP) expressions are derived and evaluated. This is the first attempt to analytically evaluate the BER in NOMA systems. The expressions are evaluated in order to propose the best combination of MCS and PAF. The trade-offs involved in selecting a certain combination of transmit constellations and PAF are also investigated. Moreover, numerical evaluations prove that the attainable BEP at the near user is affected by that of the far user. Look-up tables of suitable combinations of PAF and transmit constellations are derived using the BEP expressions, with a focus on enhancing the system data rate for some given BER constraints at both users.