Classification-based Optimal Beamforming for NOMA Wireless Relay Networks

In this paper, we study the problem of optimal beamforming for non-orthogonal multiple-access (NOMA) wireless relay networks. For a two-hop wireless relay network that consists of a NOMA broadcasting channel and a Gaussian interference channel, we propose a novel algorithm that efficiently obtains beamforming vectors for maximizing the end-to-end sum rate. We first classifies NOMA wireless relay networks into two classes based on the channel coefficients. For a wireless relay network that belongs to the first class, we use the maximal-ratio transmission (MRT) technique to obtain an optimal set of beamforming vectors. On the other hand, for a wireless relay network that belongs to the second class, we transform the non-convex optimal beamforming problem into a semidefinite programming (SDP) problem and then solve it. Simulation results show that the proposed approach significantly outperforms a number of alternative schemes such as the random-phase beamforming scheme, the maximal-ratio transmission scheme, and the zero-forcing beamforming scheme.