Fixed-Point MIMO Detector for 5G New Radio

Abstract

Fixed-point simulation of multiple-input and multiple-output (MIMO) with ordered successive interference cancellation (OSIC) detector for 5G new radio (NR) is presented. 5G NR is in demand to provide high data rate (in Gbps) and the massive multiple devices connectivity. The above requirements are fulfilled with the incorporation of multiple antenna based transceiver in 5G NR. The spatial multiplexed MIMO system resulted from aforesaid incorporation effectively increases the data rate using MIMO detectors. The performance of various optimal and sub-optimal MIMO detectors used for MIMO channel matrix inversion is limited to the singularity of the channel matrix. While triangularization and orthogonalization are prerequisites for solving the singularity of the channel matrix in MIMO systems, QR decomposition using Gram-Schmidt (GS) method is one of the stable technique. Also, using the QROSIC detection, the system succeeds to achieve near maximum likelihood (ML) performance. In this work, we investigated the fixed-point implementation of QR-OSIC MIMO detector over 5G NR. The simulation results also validate the proposed fixed-point detector succeed to achieve near ML performance while maintaining the word precision low compared to floating-point implementation.