<p>Although the mean square error (MMSE) approach is recognized to be near optimal for uplinking large-scale multiple-input-multiple-output (MIMO) systems, there are certain difficulties in the procedure related to matrix inversion. The long recurrence enlarged conjugate gradient (LRE-CG) approach is proposed in this study as a way to iteratively realize the MMMS algorithm while avoiding the complications of matrix inversion. In addition, a diagonal-approximate starting solution to the LRE-CG

more »

... proach was used to speed up the conversion rate and reduce the complications required. It has been discovered that the LRE-CG-based approach has the ability to significantly reduce computational complexity. By comparing simulation results, it is clear that this new methodology surpasses well-established wayslike the Neumann series approximation-based method and the Gauss-Siedel iterative method. With a small number of iterations, the suggested approach achieves near-optimal performance of a standard MMSE algorithm.</p>