Data Detectors for Massive MIMO Systems Using Enhanced Acceleration Overrelaxation and Special Matrices Structures

Abstract

Massive multiple-input multiple-output (MIMO) is a cornerstone technology in beyond 5G (B5G) communication systems due to its ability to achieve exceptional power and spectral efficiency. The development of low-complexity detectors for massive MIMO remains a key area of research, driven by the need to strike a balance between performance and computational complexity, especially as the number of antennas increases at both the transmitter and receiver. In this paper, we propose efficient initialization methods to address these challenges. Instead of the conventional diagonal matrix, we employ the stair matrix and the band matrix in the initialization of the proposed detector based on accelerated overrelaxation. We also employ successive overrelaxation, Gauss-Seidel, and Jacobi methods to improve the performance of the proposed detector. The initialization scaling factors are based on the spectral radius of the iteration matrix. The proposed detectors are evaluated using diverse massive MIMO configurations and multiple modulation schemes and under both perfect and imperfect channel state information (CSI). Extensive simulations show that the proposed detectors achieve significant performance enhancements accompanied by a remarkable reduction in computational complexity, making them highly suitable for practical large-scale systems.