Iterative Nonlinear Detection and Decoding in Multi-User Massive MIMO

Abstract

In this paper, we propose a new candidates list re-calculating to improve performance of iterative nonlinear detection and decoding in Multi-User (MU) Massive Multiple Input, Multiple Output (MIMO) system. The proposed nonlinear iterative detector includes a new algorithm of users (UEs) sorting before QR decomposition (QRD) and a new sorting-reduced (SR) K-best method. If MIMO detector is based on a candidates list updates, the performance can be improved by the candidates list re-calculating or using a priori information in the list generation. This is natural, because the quality of the candidates list is likely to be improved by using the decoder output as a priori information. We analyze the convergence of combining the detection algorithms with the soft low-density parity-check (LDPC) decoder. Simulation results are presented in 5G QuaDRiGa channel with QAM64 modulation in 48 × 64 MIMO system and compared with other state-of-art approaches.