Performance analysis and comparison of approximate detectors with linear detectors in massive MIMO systems under imperfect channel state information

Massive MIMO systems are a key technology for enhancing the capacity and reliability of wireless communication systems. However, to unlock the full potential of massive MIMO systems, accurate channel state information is crucial at the receivers. In practice, channel estimation errors are inevitable, leading to significant performance degradation in traditional massive MIMO equalizers that assume perfect channel state information. In this work, we investigate the performance of approximate massive MIMO detectors, specifically the Gauss-Seidel and Neumann series, over a time-invariant channel with imperfect channel state information. We derive the expressions of signal-to-interference-plus-noise ratio and average bit error rate for approximate massive MIMO detectors in a unicellular environment using M-ary quadrature amplitude modulation. Additionally, we compare the performance of linear detectors, including zero-forcing, minimum-mean-squared-error, and maximal-ratio combining, with that of the approximate detectors. Through extensive Monte Carlo numerical simulations and various system configurations, we demonstrate that the Gauss-Seidel detector, under imperfect channel state information, achieves comparable bit error rate performance to the zero-forcing and minimum-mean-squared-error detectors while requiring less computational complexity. Our study suggests that the Gauss-Seidel detector has the potential to be used in massive MIMO systems in the presence of imperfect channel state information, highlighting its practical relevance in real-world scenarios.