Parallel Group Detection Approach for Massive MIMO systems

In this paper, we develop a new detection algorithm, called Parallel Group Detection (PGD), which enables the classical detectors to improve Bit Error Rate (BER) performance of Massive MIMO systems. In the algorithm, the Massive MIMO system equation is converted into an extended form and divided to provide two sub-systems, which can subsequently be detected by the classical detectors to recover transmitted signals. Based on the PGD, the conventional QRD and SQRD detectors, we propose two efficient detectors, called QRD based PGD (QRD-PGD) and Sorted QRD based PGD (SQRD-PGD). Numerical and simulation results for different antenna configurations show that the proposed detectors can achieve a gain of more than 4 dB in BER performance as compared to that of the Minimum Mean Square Error (MMSE) one while their computational costs are comparable.