New User Grouping Scheme for Better User Pairing in NOMA Systems

Non-orthogonal multiple access (NOMA) guarantees good spectral efficiency and better enhance system radio capacity. However, the deployment of NOMA is not easy due to complexity of successive interference cancellation (SIC). Thus, it is advisable to divide users into orthogonal frequency groups and adopt a Conventional NOMA (C-NOMA) pairing within each group (hybrid NOMA). Knowing that C-NOMA uses the channel gain difference and the power allocation to increase the data rate of the pair, in a dense user environment, this channel gain difference narrows further. To do this, the users are grouped in such a way as to exploit the intergroup gap in order not only to avoid the formation of very small channel gain pairs, but also improved the data throughput of low gain groups. In this paper, we examine the user pairing scheme that we will call D-NOMA (Distributed NOMA), in a uniform deployment scenario, to distribute the data throughput in the NOMA sets, to exploit the intergroup gap and facilitate a controlled data throughput distribution of NOMA weak pairs more precisely. The results show that about 50% of D-NOMA pairs of small channel gain difference benefit from a data rate higher than C-NOMA, and 25% of D-NOMA pairs of small channel gain difference benefit from a higher data rate than conventional Divide-and-Next-Largest-difference based User Pairing Algorithm (D-NLUPA).