A Preconditioned Symmetric Successive Over Relaxation Method for Massive MIMO Signal Detection

Abstract

For a given massive multiple-input multiple-output (MIMO) systems, how to achieve near-optimal performance with low system complexity is crucial for practical usage. To solve this problem, a new preconditioned symmetric successive over relaxation (SSOR) for minimum mean square error (MMSE) method is introduced in this paper, and the simulation results are compared with other common approaches. The preconditioned process is based on NS method which is found to be able to produce accurate enough initial guess that can be used to improve the convergence rate in the subsequent SSOR processing. The simulation results show that the combined system is improved in convergence rate with limited affect in complexity and bit error rate (BER).