Performance-Complexity Tradeoffs of MIMO-NOMA Receivers Towards Green Wireless Networks

Abstract

In this paper, new layered multiuser detection (L-MUD) designs for the uplink MIMO-NOMA are investigated to optimize the tradeoffs between achieved performance and complexity. Linear/nonlinear based MUD technique is used for each multiuser MIMO (MU-MIMO) group and a successive interference cancellation process is used for the considered NOMA layers. The upper bound of symbol error probability is derived for MIMO-NOMA with receive diversity over Rayleigh fading channel to show the impact of proposed L-MUD schemes on the achieved performance. Compared with the existing minimum mean squared error (MMSE) based L-MUD and the generic MU-MIMO receivers, numerical simulations demonstrate valuable tradeoffs between the permissible user overloading, required number of radio frequency chains (RFCs), computational efforts, and bit error rate. The new outcomes of complexity reductions will lead to significant power saving and extend the state-of-the-art of MIMO-NOMA towards green applications for 5G wireless networks and beyond.