On the Performance of Lattice Reduction Schemes for MIMO Data Detection

Abstract

Recently the use of lattice reduction (LR) methods for data detection in multiple-input multiple-output (MIMO) systems has been proposed in order to achieve full diversity with suboptimal detection schemes. To this end, several reduction criteria and algorithms known from lattice theory have been applied. In this work new insights about the applicability of the various LR methods to linear and non-linear detection schemes are developed. In fact, the crucial performance parameters for linear and non-linear schemes are their associated post- equalization SNRs. It turns out that Seysens's LR algorithm and LLL reduction perform differently with respect to these post-equalization SNRs, which explains their different error-rate behavior in context of linear as well as non-linear data detection.