Enhanced Hybrid Detection Technique for Minimum Mean Square Equalizer in Uplink Massive MIMO Systems

Abstract

Uplink massive Multiple-Input Multiple-Output (MIMO) systems include huge antennas in the base station (BS) that simultaneously serve fewer Users with single-antenna, making signal detection a major issue Due to the huge matrix inversion requirement. We propose an iterative detection technique based on the enhanced Alternating Direction Method of Multipliers–Conjugate Gradients (ADMM-CG) to avoid direct matrix inversion. The ADMM is first applied as an initial vector. Then, the CG iteration algorithm terminates the calculations for the rest of the iterations. A low-complexity initial method based on trace tridiagonal has been proposed to improve the suggested technique’s performance. This integration is crucial for optimizing the tradeoff between performance and complexity. The proposed technique outperforms traditional iterative approaches regarding signal detection performance. Monte Carlo simulations show that the proposed detector performs near optimally, decreases complexity, and necessitates fewer iterations.Furthermore, it outperforms existing solutions, which are sensitive at high modulation orders and when the number of users approaches the entire number of antennas in the base station.

Keywords