Adaptive MSER Reduced-Rank Decision Feedback Receiver for Large-Scale Multiple-Antenna Systems

In this paper, we propose a novel adaptive reducedrank decision feedback (DF) receiver strategy based on the minimization symbol-error-rate (MSER) criterion for largescale multiple-antenna systems. In this strategy, a reduced-rank method for general parameter estimation using an adaptive joint and iterative preprocessing, decimation and filtering (JPDF) scheme is proposed for DF receive processing. We design a pre-stored decimation matrix and develop conjugate gradient (CG) and modified conjugate gradient (MCG) algorithms for computing the parameters of the preprocessor and reduced-rank filter. A computational complexity analysis is carried out for the existing and proposed adaptive DF algorithms. Simulation results are provided and show that the proposed adaptive MSER-JPDF reduced-rank DF receive processing strategy can achieve a superior performance to conventional adaptive linear and DF receive processing algorithms.