An improvement of ZF/LR method in MIMO detection

Abstract

In this paper, we propose a simple method which may enhance the performance of zero-forcing (ZF) or lattice reduction (LR) algorithm in multiple input multiple output (MIMO) detection. As we know MIMO communication systems increase the spectral efficiency and data reliability. Designing detector for such systems that have low complexity and performance near to optimum have attracted much of attention. Optimum detection due to its complexity is not applicable. Famous suboptimal algorithms are ZF, Minimum Mean Square Error (MMSE) and LR. ZF and MMSE are known as linear detectors although are very simple, their performance are far from optimum. LR technique were proposed to improve the linear detector performances. In this paper we propose a method that can improve the performances of the linear detectors as well as LR. This method is applied to the output of the above mentioned MIMO detection algorithm (ZF, MMSE and LR) and employing this point as the initial lattice point and search the neighboring point for better metric in the lattice. Simulation results confirm the reduction of the bit error rate by this algorithm.